JP2015180982A - Foreign matter detection device and foreign matter detection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect, when paper is bound by a material made of metal and/or other than metal and is put into a paper reception part of a paper processing device, the material binding the paper as a foreign matter.SOLUTION: A foreign matter detection device receives bundled paper in a depositing and dispensing port 140; photographs an image of the received bundled paper with a camera 152; detects a straight line appearing in a direction substantially parallel to the thickness direction of the paper captured in the photographed image, in an area excluding areas capturing ribs 142a, 142b, 142c, and 142d of the photographed image, and detects a material binding the paper as a foreign matter on the basis of the detection.

Description

  The present invention relates to a case where a paper sheet bundle is inserted in a state where the paper sheet processing apparatus for processing paper sheets such as banknotes, checks, gift certificates, and vouchers is bound with a banding band and a rubber band. The present invention relates to a foreign object detection device and a foreign object detection method for detecting foreign objects such as rubber bands.

  2. Description of the Related Art Conventionally, there has been known a technique for detecting a mixed foreign object when a foreign object other than a banknote is mixed into a banknote deposit / withdrawal port of the banknote depositing / dispensing apparatus. For example, in Patent Document 1, when an image of a foreign object receiver provided at the lower part of a bill insertion slot is captured and the image of the foreign object receiver is compared with an image without a foreign object, a foreign object is mixed. A technique is disclosed in which an image of a captured foreign object is displayed and whether or not it is necessary to return is confirmed with a user.

  However, when the foreign matter put into the banknote deposit / withdrawal port does not fall into the foreign matter receptacle below the banknote deposit / withdrawal port, the foreign matter cannot be detected. For example, when a bill bundle is inserted into a banknote deposit / withdrawal port in a state of being bundled with a money clip, the money clip cannot be detected by the technique disclosed in Patent Document 1 because it does not fall into the foreign object receptacle.

  As described above, there is also known a technique that can detect a foreign object when a bundle of money bundled with a money clip is inserted into a bill deposit / withdrawal port. For example, Patent Document 2 discloses a technique in which a magnetic sensor is provided on a paper sheet mounting table of a paper sheet feeding device and metal foreign objects are detected based on a change in the magnetic level of the magnetic sensor.

JP 2007-34936 A Japanese Utility Model Publication No. 2-149649

  However, when a bundle of banknotes bundled with non-metal rubber bands or band seals is inserted into the banknote deposit / withdrawal slot, not only does it fall into the foreign object receptacle under the banknote deposit / withdrawal slot, Therefore, it is not detected by a magnetic sensor. Therefore, when a bill bundle in a state of being bound with a rubber band or a band seal is inserted into the banknote deposit / withdrawal port, the banknote deposit / withdrawal apparatus starts a payout process to perform authenticity determination and counting of the inserted banknote, but is bundled. The machine cannot be paid out normally because of the status. Moreover, since the banknote depositing / withdrawing apparatus cannot recognize that the banknotes are bound, the banknote depositing / dispensing apparatus cannot tell the user the cause of the feeding error. Moreover, the banknote may be damaged by trying to forcibly feed the banknote in a bound state.

  For this reason, in a banknote depositing and dispensing device such as ATM (Automated Teller Machine), when a banknote bundle bundled by a non-metal rubber band or a band seal is inserted into a banknote deposit / withdrawal port, the rubber band or band seal is bound. It is an important issue to detect an object as a foreign object.

  The present invention is for solving the above-described problems of the prior art, and is a paper sheet processing apparatus in a state in which the paper sheets are bound by some binding material including a metal and / or a material other than the metal. It is an object of the present invention to provide a foreign object detection device and a foreign object detection method that can detect a bundled object as a foreign object when it is inserted into the receiving unit.

  In order to solve the above-described problems and achieve the object, the present invention is a foreign object detection device for detecting a foreign substance thrown together with an accepted paper sheet, and is a paper sheet receiving means for receiving a bundle of paper sheets An image capturing unit that captures an image of the paper sheet in a bundle state from a direction in which the thickness of the bundle can be confirmed, and a target region specifying unit that specifies a region to be detected by a foreign object from the image captured by the image capturing unit. Edge detection means for detecting an edge in a direction perpendicular to the thickness direction of the bundle of paper sheets from the image of the foreign substance detection target area specified by the target area specifying means, and detection by the edge detection means The number of pixels connected in a straight line of the edge image composed of the edges is counted, and when there are more than a predetermined number of pixels, a straight line detection means for detecting a straight line from the edge image, and the straight line detection. Based on the detected straight line by means, characterized in that the binding material for binding the sheet and a foreign object detecting means for detecting a foreign substance.

  Further, the present invention is the above invention, wherein the target area specifying unit is configured such that when there is an opaque member between a bundle of paper sheets received by the paper sheet receiving unit and the image capturing unit, the image An area for an opaque member is excluded from the image captured by the imaging means.

  Further, the present invention is characterized in that, in the above invention, the paper sheet receiving means is transparent in a member existing between the received paper sheets in a bundled state and the image capturing means.

  Further, the present invention is characterized in that, in the above-mentioned invention, the image capturing means captures an RGB color image, and detects foreign matter based on the captured R image, G image, and B image.

  Further, the present invention is the above invention, wherein the edge detection unit extracts the saturation from the image captured by the image capturing unit in the foreign object detection target region specified by the target region specifying unit. An image or a brightness image from which brightness is extracted is generated, and edge extraction is performed using the saturation image and / or the brightness image.

  Further, the present invention is the above invention, further comprising: a light source that emits light in a single or a plurality of first directions; and a light source that emits light in a second direction, each light source receiving the paper sheet Is disposed at different positions with respect to the direction perpendicular to the thickness direction of the bundle of paper sheets received by the means, and the image imaging means irradiates the surface of the bundle of paper sheets to be imaged in the two directions. And a light irradiation control means for controlling light emission of the light source.

  Further, the present invention is the above invention, wherein the foreign object detection means turns on the light source that emits light in the first direction by the light irradiation control means and irradiates the paper with light. And the second image captured by the image capturing unit in a state where the light irradiation control unit turns on the light source that emits light in the second direction and irradiates the paper with light. A foreign object is detected based on an evaluation of a difference between the first image and the second image in an area including a straight line detected by the straight line detection unit. To do.

  Further, the present invention is the above invention, wherein the image capturing unit is capable of acquiring a high resolution image and a low resolution image from the captured image, and the straight line detecting unit is acquired by the image capturing unit. When a straight line is detected based on a low-resolution image, straight line detection processing is performed based on the high-resolution image acquired by the image capturing unit.

  Further, in the present invention according to the above invention, the foreign object detection unit is sandwiched between the two straight lines of the image captured by the image capturing unit, and the distance between the two straight lines detected by the straight line detection unit is a predetermined value or more. It is characterized in that foreign matter is detected when the pixel value of the target area is within a predetermined range.

  Further, according to the present invention, in the above invention, a thickness detection unit that detects a thickness of the bundle of sheets received by the sheet reception unit, and the bundle of sheets received by the sheet reception unit. A position changing unit that changes a positional relationship between the leaf and the light source, and the image capturing unit has a bundle thickness detected by the thickness detecting unit equal to or greater than a predetermined value, When a straight line is detected by the straight line detection means, the positional change means changes the positional relationship between the bundled paper sheets and the light source, and then re-captures the image of the bundled paper sheets. It is characterized by.

  Further, the present invention is a foreign matter detection method for detecting a foreign matter thrown together with a received paper sheet, from a paper sheet receiving step for receiving a bundle of paper sheets, and a direction in which the thickness of the bundle can be confirmed. An image capturing step for capturing an image of the paper sheet in a bundled state, a target region specifying step for specifying a target region for detecting foreign matter from the image captured by the image capturing step, and a target region specifying step An edge detection step for detecting an edge in a direction perpendicular to the thickness direction of the bundle of paper sheets from the image of the foreign object detection target region, and a linear shape of an edge image composed of the edges detected by the edge detection step And when there are more than a predetermined number of pixels, a straight line detecting step for detecting a straight line from the edge image, and the straight line detecting step. Tsu based on straight lines detected by the flop, characterized in that the binding material for binding the sheet containing a foreign matter detection step of detecting a foreign body.

  The present invention is also a foreign object detection method for detecting a bandage wound around a loaded paper sheet as a foreign object, the paper sheet receiving step for receiving a bundle of paper sheets, and the thickness of the bundle An image capturing step for capturing an image of the paper sheet in a bundled state from a direction that can be confirmed, a target region specifying step for specifying a target region for detecting foreign matter from the image captured by the image capturing step, and the target region specifying step An edge detection step for detecting an edge in a direction perpendicular to the thickness direction of the bundle of paper sheets from the image of the region to be detected of the foreign matter specified by the edge, and an edge constituted by the edge detected by the edge detection step The number of pixels connected to a straight line of the image is counted, and when there are more than a predetermined number of pixels, a straight line detecting step for detecting a straight line from the edge image, and the straight line In the exiting step, when a straight line is detected within a predetermined distance from the edge of the paper sheet, the image of the paper sheet captured by the image capturing step is continuously from the edge of the paper sheet. A mask region determination step for determining, as a mask region, a region sandwiched between the edge of the paper sheet and the straight line detected in the straight line detection step when there is a region that maintains uniform brightness with uniformity; Detect pixels that are brighter than the pixels in the thickness direction of the bundle of paper sheets in the area other than the mask area determined in the mask area determination step and that are brighter than the pixels in the thickness direction of the bundled paper sheet The paper sheet is bound based on the shining pixel detecting step, the straight line detected by the straight line detecting step, and the shining pixel detected by the shining pixel detecting step. Characterized in that it includes a foreign matter detection step of detecting the seal as foreign.

  Further, the present invention is the above invention, wherein the straight line detecting step specifies a start point and an end point indicating a range in which the effective edge image that is the basis of the straight line and the straight line exists, and the mask region determining step includes The mask for each straight line, which is a region sandwiched between the edge of the paper sheet and the straight line among the thickness direction regions sandwiched by the position in the thickness direction of the paper sheet corresponding to the start point and the end point with respect to the straight line When the area is determined and a plurality of the straight lines are detected within a predetermined distance from the edge of the paper sheet, the thickness direction areas of the straight line mask areas corresponding to the respective straight lines overlap. The overlapping mask region corresponding to the thickness direction region is defined as a region corresponding to the straight line mask region of the straight line closest to the edge of the paper sheet, and the straight lines corresponding to the straight lines. When the thickness direction area mask regions do not overlap, the mask area with respect to the thickness direction region not overlapping, characterized by all of the straight lines superimposed said linear specific mask region corresponding region.

  According to the present invention, a paper sheet in a bundle state is received, an image of the paper sheet in a bundle state is taken from a direction in which the thickness of the bundle can be confirmed, a region for detecting a foreign object is identified from the taken image, Edges in the direction perpendicular to the thickness direction of the bundle of paper sheets are detected from the image of the specified foreign object detection target area, and the number of pixels connected in a straight line of the edge image composed of the detected edges is counted and predetermined. Since a straight line is detected when there are more than a few pixels, and a binding object that binds paper sheets based on the detected straight line is detected as a foreign object, the paper sheets are made of metal and / or metal other than metal. When the sheet is fed into a paper sheet accepting unit of the paper sheet processing apparatus in a state of being bound by some bundle including the material, the bundle can be detected as a foreign object.

FIG. 1 is an explanatory diagram for explaining the outline of the foreign object detection method according to the first embodiment. FIG. 2 is an explanatory diagram for explaining the details of the appearance and the internal physical configuration of the deposit / withdrawal port shown in FIG. 1. FIG. 3 is a block diagram showing an internal configuration of the banknote depositing and dispensing machine according to the first embodiment shown in FIG. FIG. 4 is a flowchart showing a processing procedure during the depositing process of the banknote depositing and dispensing machine shown in FIG. FIG. 5 is a block diagram showing an internal configuration of the deposit / withdrawal port foreign matter detection unit of the banknote depositing / dispensing machine shown in FIG. 3. FIG. 6 is a flowchart showing the entire processing procedure of foreign object detection processing by the deposit / withdrawal port foreign object detection unit shown in FIG. FIG. 7 is an explanatory diagram for explaining the processing content of the foreign object detection target region extraction processing shown in FIG. FIG. 8 is a flowchart showing a detailed processing procedure of the straight line detection processing shown in FIG. FIG. 9 is a flowchart showing a detailed processing procedure of the continuity evaluation process shown in FIG. FIG. 10 is an explanatory diagram for explaining the processing content of the continuity evaluation processing shown in FIG. FIG. 11 is an explanatory diagram for explaining the processing content of the bundling determination processing shown in FIG. 6. FIG. 12 is an explanatory diagram for describing control of the light source unit related to the three-dimensional object determination process illustrated in FIG. 6. FIG. 13 is a flowchart illustrating a detailed processing procedure of the three-dimensional object determination process illustrated in FIG. 6. FIG. 14 is a diagram for explaining an outline of a more strict banding determination process according to the second embodiment. FIG. 15 is a view for explaining the internal structure and operation of the deposit / withdrawal port shown in FIG. 2 according to the second embodiment. FIG. 16 is a diagram for explaining the outline of the method for specifying the shining pixel of the banknote bundle according to the second embodiment. FIG. 17 illustrates a mask range that is a non-evaluation area that does not search for a shining pixel used for the band seal determination when a bank note that has a white area at the edge of the bank note illustrated in FIG. It is a figure for doing. FIG. 18 is a flowchart illustrating an entire processing procedure of foreign object detection processing by the deposit / withdrawal port foreign object detection unit according to the second embodiment.

  Exemplary embodiments of a foreign object detection device and a foreign object detection method according to the present invention will be described below in detail with reference to the accompanying drawings. In the embodiment, description will be made using “banknotes” as a representative of “paper sheets”.

  First, an outline of the foreign object detection method according to the first embodiment will be described. FIG. 1 is an explanatory diagram for explaining the outline of the foreign object detection method according to the first embodiment. FIG. 1A is an external view of the banknote depositing / dispensing machine 100. FIG.1 (b) is a figure which shows the outline | summary of the physical structure of the deposit or withdrawal port (paper sheet reception part) 140 which inserts a banknote at the time of the depositing of the banknote depositing / withdrawing machine 100 shown to Fig.1 (a). FIG. 1C is an example of an image obtained by capturing the banknotes inserted into the deposit / withdrawal port 140 by the camera 152 shown in FIG. FIG. 1D is a diagram for explaining a flow of processing for detecting, as a foreign object, a rubber band, a band seal, or the like that binds banknotes using the image shown in FIG.

  The external structure of the banknote depositing / dispensing machine 100 will be described with reference to FIG. The display operation unit 110 is a touch panel display device, and can accept input by displaying a menu, a transaction screen, a message to the user, and operating the displayed menu and a button on the transaction screen. The card insertion slot 121 is a slot into which a magnetic card such as a cash card or an IC card is inserted. The number input unit 130 is a keyboard that mainly inputs numbers such as a password and an amount. The deposit / withdrawal port 140 is a port into which a banknote to be deposited is inserted when performing a deposit-type transaction, and is a port into which a banknote to be withdrawn is dispensed when performing a withdrawal-type transaction. The statement issuance port 191 is a port for discharging a transaction statement on which transaction details are printed. The voice call unit 200 is an input / output unit for a user to make a voice call with a back office staff when using the banknote depositing and dispensing machine 100.

  In the first embodiment, the banknotes inserted into the deposit / withdrawal port 140 when the bundle of banknotes inserted into the deposit / withdrawal port 140 of the banknote depositing / dispensing machine 100 shown in FIG. An example in which the rubber band or band seal is detected as a foreign object before the bundle feeding operation will be described. By detecting the rubber band or the band seal as a foreign object, it is possible to notify the user of a message indicating a re-insertion instruction after removing the rubber band or the band seal.

  With reference to FIG. 1B, the internal physical configuration of the deposit / withdrawal port 140 will be described. A shutter 141 is provided above the deposit / withdrawal port 140 and opens when a deposit transaction operation is performed, and closes when a bill to be deposited into the deposit / withdrawal port 140 is inserted. In addition, there are ribs 142a, 142b, 142c, and 142d at the bottom of the deposit / withdrawal port 140. For example, when a foreign object such as a coin is accidentally inserted, And can be separated. In addition, a camera 152 is installed at the bottom of the deposit / withdrawal port 140 below the ribs 142a, 142b, 142c, 142d, and the camera 152 captures an image of a bill inserted into the deposit / withdrawal port 140 from below. Can do.

  FIG.1 (c) is an example of the image of the banknote imaged with the camera 152 of FIG.1 (b). Since the ribs 142a, 142b, 142c, and 142d exist between the camera 152 and the banknote bundle, the ribs 142a, 142b, 142c, and 142d are reflected in the three image examples. The top image is an example of an image of banknotes that are not bound by a rubber band or a bandage, the middle image is an example of an image of a banknote that is bound by a rubber band, and the bottom image is a banknote that is bound by a bandage. It is an example of an image.

  The rubber band image is reflected in the middle image in FIG. In addition, a banded image is reflected in the bottom image in FIG. The first embodiment will explain an implementation example in which a bundle of rubber bands, bandages, or the like reflected in these images is detected as a foreign object. As shown in FIG. 1C, the foreign matter binding the banknotes has a feature that it appears as a linear shape that is substantially parallel to the thickness direction of the banknotes on the image. By analyzing the obtained image, foreign matter that binds banknotes such as rubber bands and bandages is detected. However, since the ribs 142a, 142b, 142c, and 142d are fixedly reflected due to the physical structure of the deposit / withdrawal port 140, the portion where the ribs 142a, 142b, 142c, and 142d are reflected is a region that is not a foreign object detection target. By doing so, the reflection of the ribs 142a, 142b, 142c, 142d is not erroneously detected as a foreign object.

  With reference to FIG. 1 (d), a flow of processing related to foreign object detection for binding banknotes by the banknote depositing and dispensing machine 100 will be described. If the user accepts the insertion of a bill into the deposit / withdrawal port 140 in the deposit transaction, the bill deposit / withdrawal machine 100 closes the shutter 141 and captures an image of the bill inserted by the camera 152.

  The banknote depositing / dispensing machine 100 analyzes the image captured by the camera 152 to determine whether or not there is a foreign object that binds banknotes such as a rubber band or a band seal. A message prompting re-insertion is displayed after removing the foreign matter being opened, and the shutter 141 is opened to wait for re-input. In addition, when it is determined from the image captured by the camera 152 that there is no foreign object that binds the banknote bundle, a feeding process for continuing the depositing process of the banknote inserted into the deposit / withdrawal port 140 is started.

  In this way, the banknotes in a bundled state are received by the deposit / withdrawal port 140, the image of the accepted banknotes in the bundled state is captured by the camera 152, and the areas where the ribs 142a, 142b, 142c, 142d of the captured images are reflected are excluded. The foreign matter added to the banknote is detected based on the detection of the straight line appearing in the direction substantially parallel to the thickness direction of the banknote in the captured image. When inserted into the depositing / dispensing port 140 of the banknote depositing / dispensing machine 100 in a state of being bound by some binding object including a material other than metal, it is possible to detect a foreign substance binding the banknotes.

  Next, the details of the external appearance and the internal physical configuration of the deposit / withdrawal port 140 shown in FIG. 1 will be described. FIG. 2 is an explanatory diagram for explaining the details of the appearance of the deposit / withdrawal port 140 and the internal physical configuration. FIG. 2A is a detailed external view of the deposit / withdrawal port 140. FIG. 2B is a view for explaining the structure of the portion below the ribs 142a, 142b, 142c, 142d of the deposit / withdrawal port 140. FIG. FIG. 2C is a diagram for explaining the arrangement of the LEDs 151a to 151k, the ribs 142a, 142b, 142c, 142d, and the camera 152 of the light source unit 151 arranged under the ribs 142a, 142b, 142c, 142d. .

  An external configuration of the deposit / withdrawal port 140 will be described with reference to FIG. A shutter 141 is provided above the deposit / withdrawal port 140 and is closed in a standby state, and is opened when a bill for a deposit transaction or a bill for a withdrawal transaction is thrown out. The banknote holder 143 is an operation unit that presses a banknote inserted after the banknote is inserted into the deposit / withdrawal port 140 and the shutter 141 is closed in a deposit transaction. In addition, there are ribs 142a, 142b, 142c, 142d below the deposit / withdrawal port 140, and the inserted bills are placed on the ribs 142a, 142b, 142c, 142d. Also. A camera 152 is disposed further below the ribs 142a, 142b, 142c, 142d, and images the banknotes placed on the ribs 142a, 142b, 142c, 142d.

  As shown in FIG. 2B, the camera 152 and the light source unit 151 are disposed below the ribs 142a, 142b, 142c, and 142d of the deposit / withdrawal port 140. When the camera 152 takes an image, the light source unit 151 is turned on. Then, the imaging object is irradiated with light.

  With reference to FIG. 2C, the positional relationship between the LEDs 151a to 151k of the light source unit 151, the ribs 142a, 142b, 142c, 142d, and the camera 152 will be described. The light source unit 151 is placed below the ribs 142a, 142b, 142c, 142d and placed in order to capture the bills placed on the ribs 142a, 142b, 142c, 142d with the camera 152 from below. Irradiate light on the lower surface. Further, as shown in FIG. 2C, when the entire area is divided into five areas at the positions of the ribs 142a, 142b, 142c, 142d, LEDs are arranged at both ends of each area and in the middle thereof. . By adopting such an LED arrangement, it is possible to minimize the influence of the shadows of the ribs 142a, 142b, 142c, 142d, and to detect a three-dimensional object by comparing images picked up by illuminating light sources at different positions. It is possible to do. Detailed lighting control of the LEDs 151a to 151k for detecting the three-dimensional object will be described later.

  Although not shown in the physical configuration of FIG. 2, the ribs 142a and 142b of the inserted banknotes are considered in consideration of the possibility that foreign matters such as bandages and rubber bands may overlap the ribs 142a, 142b, 142c, and 142d. , 142c, 142d may be provided with a second banknote presser that can change the position with respect to the direction in which they are arranged. By using the second banknote press to move the position of the banknote ribs 142a, 142b, 142c, 142d in the direction in which the banknotes are arranged, the same processing is performed, so that foreign matters such as band seals and rubber bands are ribbed 142a, 142b, 142c. , 142d may be taken as a countermeasure.

  Next, the internal structure of the banknote depositing / dispensing machine 100 according to the first embodiment shown in FIG. 1 will be described. FIG. 3 is a block diagram illustrating an internal configuration of the banknote depositing and dispensing machine 100 according to the first embodiment.

  The bill depositing / dispensing machine 100 includes a display operation unit 110, a card reader 120, a number input unit 130, a deposit / withdrawal port 140, a deposit / withdrawal port foreign matter detection unit 150, a transport unit 160, a discrimination unit 170, a bill storage unit 180, and a specification printer. 190, a voice call unit 200, a communication unit 210, a storage unit 220, and a control unit 230. Description of the parts already described so far will be omitted, and description will be made centering on parts that have not been described yet.

  The card reader 120 is a reading unit that reads the contents of a magnetic card such as a cash card inserted from the card insertion slot 121 or an IC card. The deposit / withdrawal port 140 includes the shutter 141, the rib 142, the bill presser 143, and the feeding unit 144 described above. The feeding unit 144 is a processing unit that feeds the bill inserted into the deposit / withdrawal port 140 to the transport unit 160 when it is confirmed that there is no foreign matter in the processing for detecting foreign matters such as rubber bands and band seals.

  The deposit / withdrawal port foreign matter detection unit 150 is a processing unit that determines whether or not a foreign matter has been added to the bill inserted into the deposit / withdrawal port 140. The conveyance part 160, the banknote paid out from the deposit / withdrawal port 140 in the deposit transaction is conveyed to the discrimination unit 170 or the banknote storage unit 180, or the banknote in the banknote storage unit 180 is conveyed to the deposit / withdrawal port 140 in the withdrawal transaction. It is a processing unit.

  The discrimination unit 170 is a processing unit that determines the type of banknotes that have been transported by the transport unit 160 and authenticity determination. The banknote storage unit 180 is a storage unit that stores banknotes that have been inserted in deposit transactions and determined to be genuine bills, and a storage unit that stores banknotes for dispensing banknotes in withdrawal transactions. is there. The specification printer 190 delivers the specification to the user by discharging it to the specification issuing port 191 with an output unit for printing the specification of the transaction for the user. The communication unit 210 is an interface unit for data communication with a system for managing account information related to deposit / withdrawal transactions and deposits / withdrawals.

  The storage unit 220 is a storage device such as a hard disk device or a non-volatile memory, and stores data necessary for deposit / withdrawal processing. The control unit 230 is a control unit that controls the entire bill depositing / dispensing machine 100, and includes a deposit processing unit 231 and a withdrawal processing unit 232. Actually, it is realized by storing programs corresponding to these functional units in a non-illustrated non-volatile memory such as a ROM, and loading and executing these programs on a CPU (Central Processing Unit).

  The deposit processing unit 231 is a control unit that controls the entire deposit process performed by the banknote depositing and dispensing machine 100, and the withdrawal process unit 232 is a control unit that controls the entire deposit process performed by the banknote depositing and dispensing machine 100. . A detailed processing flow at the time of deposit processing will be described later.

  Next, a processing procedure at the time of depositing processing of the banknote depositing / dispensing machine 100 shown in FIG. 1 will be described. FIG. 4 is a flowchart showing a processing procedure during the depositing process of the banknote depositing / dispensing machine 100 shown in FIG.

  The deposit processing unit 231 accepts a card from the card insertion slot 121, reads the contents of the card inserted by the card reader 120, and accepts execution of a deposit transaction by the display operation unit 110, and then deposits / deposits the banknote to be deposited. The shutter 141 is opened so that it can be inserted into the metal cap 140 (step S101). The deposit processing unit 231 waits for a bill to be inserted into the deposit / withdrawal port 140, performs a bill insertion detection process (step S102), and if no bill is inserted into the deposit / withdrawal port 140 (step S102). In step S103; No), the process returns to step S102.

  When the insertion of the bill into the deposit / withdrawal port 140 is detected (step S103; Yes), the deposit processing unit 231 closes the shutter 141 (step S104), and the deposit / withdrawal port foreign matter detection unit 150 Foreign matter detection processing is performed to determine whether or not a rubber band or banding foreign matter has been added to the banknote bundle inserted into the deposit / withdrawal port 140 (step S105).

  If no foreign matter is detected in the foreign matter detection process in step S105 (step S106; No), the deposit processing unit 231 feeds the banknote received by the feeding unit 144 of the deposit / withdrawal port 140 to the transport unit 160 (step S107). ). The banknotes fed out to the transport unit are transported to the discrimination unit 170, where the discrimination unit 170 determines the type of banknote and the authenticity of the banknote (step S108), and banknotes are stored for banknotes determined to be genuine. It is stored in the section 180 (step S109), and a statement describing the deposit transaction such as the deposit amount is issued by the statement printer 190 (step S110), and the process is terminated.

  When a foreign object is detected in the foreign object detection process in step S105 (step S106; Yes), the deposit processing unit 231 displays a foreign object added to the bill inserted into the deposit / withdrawal port 140 on the display operation unit 110. A message prompting re-insertion of the banknote is displayed on the deposit / withdrawal port 140 (step S111), and the shutter 141 of the deposit / withdrawal port 140 is opened (step S112). When the deposit processing unit 231 waits for the banknote to be re-inserted into the deposit / withdrawal port 140 and performs the detection process of the banknote insertion (step S113), and the re-insertion of the banknote into the deposit / withdrawal port 140 is not detected. In (Step S114; No), the process returns to Step S113. Moreover, when the re-insertion of the banknote into the deposit / withdrawal port 140 is detected (step S114; Yes), the process returns to step S104.

  Next, the internal configuration of the deposit / withdrawal port foreign matter detection unit 150 of the banknote depositing / dispensing machine 100 shown in FIG. 3 will be described. FIG. 5 is a block diagram showing an internal configuration of the deposit / withdrawal port foreign object detection unit 150 of the banknote deposit / withdrawal machine 100.

  The deposit / withdrawal port foreign matter detection unit 150 includes a light source unit 151, a camera 152, a communication unit 153, a storage unit 154, and a control unit 155. As described with reference to FIG. 2, the light source unit 151 and the camera 152 are disposed below the deposit / withdrawal port 140. When a bill is inserted into the deposit / withdrawal port 140, the shutter 141 is closed and the light source unit 151 irradiates the bill with light. The image of the banknote irradiated with light is captured by the camera 152. The communication unit 153 receives an instruction from the banknote depositing / dispensing machine 100 to determine the presence / absence of a foreign object, and the foreign currency determination result, an image captured by the camera 152, and the like to the banknote depositing / dispensing machine 100, which is a host device. It is an interface part to send. The storage unit 154 is a storage device such as a non-volatile memory, and stores data necessary for processing related to foreign object detection.

  The control unit 155 is a control unit that controls the entire deposit / withdrawal port foreign matter detection unit 150, and includes an automatic adjustment unit 155a, an image capturing unit 155b, a target area specifying unit 155c, a gray image generation unit 155d, an edge detection unit 155e, and a straight line detection. A unit 155f, a band seal determination unit 155g, and a three-dimensional object determination unit 155h. Actually, it is realized by storing programs corresponding to these functional units in a non-illustrated non-volatile memory such as a ROM, and loading and executing these programs on a CPU (Central Processing Unit).

  The automatic adjustment unit 155a emits light from the light source unit 151 and analyzes an image collected by the camera 152 at the time of initializing the apparatus, thereby adjusting the light amount for each LED of the light source unit 151 and adjusting various parameters of the camera 152. In addition, determination of adjustment values and storage in a memory are automatically executed. Thereby, the adjustment including the individual difference of the LED and the individual difference of the camera 152 can be automatically performed. Thereafter, using the parameter values stored in the memory, the light emission amount adjustment of the LED and the captured image are obtained so that a captured image corresponding to the purpose can be obtained.

  The image capturing unit 155 b performs lighting control of the light source unit 151 and captures an image of the banknote inserted into the deposit / withdrawal port 140 by the camera 152. Regarding image capturing, the image capturing method differs depending on the purpose of image capturing. An image to be captured for detecting a straight line is captured by emitting light from the light source unit 151 so that the overall illuminance difference is small and the influence of the shadows of the ribs 142a, 142b, 142c, and 142d is minimized. Further, when imaging is performed to detect a three-dimensional object such as a rubber band in a predetermined area, a set of images is captured with two sheets irradiated with light from different directions corresponding to the detection target area of the three-dimensional object. . The image capturing unit 155 b performs lighting control of the LEDs 151 a to 151 k of the light source unit 151 according to the purpose, captures an image with the camera 152, and registers the captured image in the storage unit 154. In addition, a low resolution image and a high resolution image can be acquired as an image for detecting a straight line. The low-resolution image is acquired by acquiring a high-resolution image captured by the camera 152 and reducing the resolution of the captured image data. Alternatively, a high-resolution image and a low-resolution image may be captured separately.

  The target area specifying unit 155c excludes areas other than banknotes such as the area where the ribs 142a, 142b, 142c, and 142d are reflected and the banknote presser 143 from the image acquired by the image capturing unit 155b. The target area for detecting the foreign matter added to is narrowed down. Detailed processing contents will be described later. The gray image generation unit 155d converts the image acquired by the image capturing unit 155b into a gray multi-tone image.

  The edge detection unit 155e is configured to detect the banknotes reflected in the captured image in the target area where the foreign object detection extracted by the target area specifying unit 155c is performed on the gray multi-tone image created by the gray image generation unit 155d. A straight line appearing in a direction substantially parallel to the thickness direction is detected. Specifically, the edge strength in the direction perpendicular to the thickness direction of the banknote in the image captured using an edge detection filter such as a Sobel filter is calculated, and pixels having an edge strength exceeding a predetermined threshold are half-cut. A straight line is detected by conversion. Details will be described later.

  The straight line detection unit 155f is a processing unit for evaluating the validity of the straight line detected by the edge detection unit 155e. Specifically, the edge detection unit 155e detects on the condition that a predetermined criterion is satisfied by evaluating the degree of gathering of edge pixels having a predetermined intensity obtained at the time of line detection on the line detected by the Hough transform. The straight line information is stored as a foreign object candidate. The straight line information detected by the edge detection unit 155e that does not satisfy the predetermined standard is excluded from the foreign substance candidates. Details will be described later.

  The bundling determination unit 155g is a processing unit that determines whether or not the straight line detected by the edge detection unit 155e and determined to be valid by the straight line detection unit 155f is due to bundling. Specifically, in many cases, the bandage has a width that is greater than a predetermined value, and a bank name or the like is printed on plain or single-color paper, and the change in luminance value is often smaller than that of a banknote portion. Whether or not the feature is detected is determined based on whether or not the feature can be detected. Details will be described later.

  The three-dimensional object determination unit 155h determines whether or not the straight line detected by the edge detection unit 155e and determined to be appropriate by the straight line detection unit 155f due to the three-dimensional object such as a rubber band. It is a processing unit for determining whether or not. Specifically, the difference between the two images irradiated with light from different directions (left and right directions) with respect to the straight line is evaluated with respect to the region where the target straight line exists, and the difference between the two images is large. In this case, it is determined that there is a three-dimensional object such as a rubber band, and when the difference is small, it is determined that the three-dimensional object does not exist in the region. Details will be described later.

  Next, the entire processing procedure of the foreign object detection process by the deposit / withdrawal port foreign object detection unit 150 shown in FIG. 3 will be described. FIG. 6 is a flowchart showing the entire processing procedure of foreign object detection processing by the deposit / withdrawal port foreign object detection unit 150.

  First, the image capturing unit 155b adjusts the amount of light of the LED so that the overall illuminance difference is reduced and lights it, captures the target image at a high resolution (step S201), and then straightens from the captured high resolution image data. Low-resolution image data for detection is generated (step S202). The target area specifying unit 155c extracts a foreign object detection target area by excluding a non-target area where foreign object detection is performed from the image data acquired by the image capturing unit 155b (step S203). The gray image generation unit 155d converts the image of the foreign object detection target region extracted in step S203 into a gray multi-tone image (step S204).

  The edge detection unit 155e detects a straight line by using Sobel transform or Hough transform on the gray multi-tone image created by the gray image generation unit 155d (step S205). If no straight line is detected in the straight line detection process of step S205 (step S206; No), it is determined that no foreign matter has been detected, and the process ends.

  If at least one straight line is detected in the straight line detection process in step S205 (step S206; Yes), it is determined whether or not the straight line detection process has already been performed using a high-resolution image (step S206). S207) When the straight line detection process has already been performed using the high-resolution image (step S207; Yes), the process proceeds to step S209. When the straight line detection process has not been performed using the high resolution image (step S207; No), the image capturing unit 155b captures the high resolution image data for the straight line detection (step S208). The process proceeds to step S203.

  When a straight line is detected in the straight line detection process in step S205 using the high-resolution image (step S207; Yes), the straight line detection unit 155f continues the straight lines detected in the straight line detection process in step S205. The sex is evaluated (step S209). If there is no straight line whose validity is confirmed by the evaluation of continuity in step S209 (step S210; No), it is determined that no foreign matter has been detected, and the process ends.

  In addition, when a straight line whose validity has been confirmed remains in the continuity evaluation in step S209 (step S210; Yes), the band seal determination unit 155g confirms the validity in the continuity evaluation process in step S209. It is determined whether or not the straight line is due to the banding (step S211). Further, the three-dimensional object determination unit 155h determines whether or not the straight line whose validity has been confirmed in the continuity evaluation process in step S209 is due to a three-dimensional object such as a rubber band (step S209). S212), the process ends.

  Next, processing contents of the foreign object detection target region extraction process (step S203) shown in FIG. 6 will be described. FIG. 7 is an explanatory diagram for explaining the processing contents of the foreign object detection target region extraction processing (step S203) shown in FIG.

  FIG. 7A shows an example of a straight line detection image captured by the image capturing unit 155b. As shown to Fig.7 (a), the rib 142a, 142b, 142c, 142d and the banknote holder 143 other than a banknote are reflected in the image imaged by the image imaging part 155b. First, the target area specifying unit 155c excludes a part corresponding to the banknote presser 143 from the image of FIG. FIG. 7B is a diagram excluding the part corresponding to the bill presser 143.

  Next, referring to FIG. 7C, a method for narrowing the area from the whole of FIG. Compared to FIG. 7B, an x-axis frequency distribution in which the number of pixels in which the brightness of the candidate area of the image in the x-axis direction is greater than or equal to a predetermined value is counted, and the brightness of the candidate area of the image in the y-axis direction is greater than or equal to a predetermined value A y-axis direction frequency distribution in which the number of pixels is counted is created. In consideration of the continuity of the banknote bundle area from the frequency distribution in the x-axis direction and the y-axis direction, the top and bottom and left and right edges in the image are determined, and the area where the banknotes are reflected is narrowed down. The region determined in this way is a region surrounded by a rectangle as shown in FIG.

  Next, with reference to FIG. 7 (d) and FIG. 7 (e), description will be given of a process for determining a mask region that is not subject to foreign object detection in the range where the banknote bundle shown in FIG. 7 (c) exists. The white area | region of FIG.7 (d) is an area | region of a banknote, and a black area | region is an area | region other than a banknote, and is a mask area | region which becomes the detection object of the foreign material which binds a banknote. In addition, since the portions where the ribs 142a, 142b, 142c, 142d are reflected and the left and right end portions of the banknote are easily misdetected as straight lines of foreign matter, the area near the arrow shown in FIG. It is a thing. Thus, with respect to the image in the rectangular range of FIG. 7C, the part other than the mask area shown in white in FIG.

  Next, a detailed processing procedure of the straight line detection process (step S205) shown in FIG. 6 will be described. FIG. 8 is a flowchart showing a detailed processing procedure of the straight line detection process (step S205) shown in FIG.

  The edge detection unit 155e calculates the edge strength in the direction perpendicular to the thickness direction of the banknote in the gray image generated by the gray image generation unit 155d using an edge detection filter such as a Sobel filter. (Step S301). Further, the edge detection unit 155e extracts a pixel (hereinafter referred to as an edge pixel) having a predetermined intensity of the image obtained by the edge detection filter (step S302).

  Further, the edge detection unit 155e votes all the edge pixels extracted in step S302 on the Hough plane (step S303). For voting on the Hough plane, in order to improve the accuracy of detecting narrow foreign objects such as rubber bands, if there is an edge pixel with a contrast gradient in the opposite direction within a certain pixel range, The weight may be increased.

  As a result of the voting process on the Hough plane in step S303, the edge detection unit 155e detects a straight line candidate by determining θ and ρ that cause the number of votes to be equal to or greater than a predetermined threshold (step S304). Moreover, since the straight line of the foreign material corresponding to what binds a banknote bundle has the characteristic detected in the direction substantially parallel to the thickness direction of a banknote, you may restrict | limit the range detected regarding (theta).

  Next, a detailed processing procedure of the continuity evaluation process (step S209) shown in FIG. 6 will be described. FIG. 9 is a flowchart showing a detailed processing procedure of the continuity evaluation process (step S209) shown in FIG.

  The straight line detection unit 155f selects one straight line from the straight lines extracted by the edge detection unit 155e (step S401), and counts the number of connected edge pixels calculated in step S302 of FIG. 8 on the selected straight line (step S401). S402). A specific description of the process of counting the number of connected edge pixels will be described later.

  If the number of connected edge pixels on the selected straight line counted in step S402 is greater than or equal to a predetermined threshold (step S403; Yes), it is determined that the straight line selected in step S401 is valid as a candidate and is a straight line candidate. Registration is performed (step S404). If the number of connected edge pixels on the selected straight line counted in step S402 is less than a predetermined threshold (step S403; No), it is determined that the straight line selected in step S401 is not valid as a candidate. It excludes from a candidate (step S405).

  Following the processing of step S404 or step S405, if there is a straight line for which the continuity evaluation has not been completed for the straight line extracted by the edge detection unit 155e (step S406; Yes), the process returns to step S401 and the unevaluated straight line Perform an evaluation. If there is no straight line for which the continuity evaluation has not been completed for the straight line extracted by the edge detection unit 155e (step S406; No), the process is terminated.

  Next, processing contents of the continuity evaluation process (step S209) shown in FIG. 6 will be described. FIG. 10 is an explanatory diagram for explaining the processing content of the continuity evaluation processing (step S209) shown in FIG.

  FIG. 10 shows the edge pixels detected in the edge detection process in step S302 in FIG. 8 in black, and shows the straight line selected in step S401 in FIG. 9 as the evaluation line. As shown in FIG. 10, there are three connected portions with adjacent edge pixels on the evaluation line.

  As shown in FIG. 10, examples of the three connecting parts are a connecting part a, a connecting part b, and a connecting part c. The connecting part a has 43 connected edge pixels, and the connecting part b has 6 connected edge pixels. The number of connection edge pixels is nine in the connection part c. Therefore, the maximum number of connected edge pixels of this evaluation line is 43.

  The evaluation of the continuity of the straight line is to evaluate the straight line detected by the Hough transform in this way by using the edge pixel of the original edge image, and the maximum number of connected edges of the connected part is predetermined. The validity of the detected straight line is determined when it is equal to or greater than the threshold. In addition, for a portion where the number of connected edges is equal to or greater than a predetermined threshold by evaluation of continuity, it is determined that the line is an effective straight line, and the start point and end point of the detected straight line are specified.

  Next, the processing content of the banding determination process (step S211) illustrated in FIG. 6 will be described. FIG. 11 is an explanatory diagram for explaining the processing content of the band seal determination processing (step S211) shown in FIG.

  FIG. 11 is an example in which the gray image generated in step S204 of FIG. 6 is displayed by superimposing the continuity evaluated straight line in step S209 of FIG. Moreover, the example of FIG. 11 is an example of an image of a portion where a bundle of banknotes is bound with a band seal. Since the bandage has a certain width, and the bank name etc. are printed on plain or single color paper, the change in the brightness value is often small compared to the banknote part. 11 are bound by a banding, as shown in FIG. 11, two straight lines corresponding to both ends of the banding are detected, the two straight lines have a certain width, or the brightness of the portion hitting the banding is flat. It is determined that it is a bandage using the fact that a characteristic such as Further, since the banding part has a feature that the edge strength in the thickness direction is smaller than that of the banknote part, it may be determined that the banding is a banding using the fact that there are few edge pixels in the thickness direction.

  Further, the graph shown below the image in FIG. 11 is a graph of frequency distribution obtained by counting the number of pixels having a predetermined range of luminance in the y-axis direction. As shown in FIG. 11, when the frequency distribution of luminance has a uniform area with a certain width, and two straight lines a and b are located at positions sandwiching the uniform luminance area, it is detected. A range between the two straight lines a and b is determined to be a bandage. Alternatively, if there is a straight line a or b in a uniform region, it may be determined that there is a banding.

  Next, control of the light source unit 151 regarding the three-dimensional object determination process (step S212) illustrated in FIG. 6 will be described. FIG. 12 is an explanatory diagram for describing control of the light source unit 151 regarding the three-dimensional object determination process (step S212) illustrated in FIG. FIG. 12A is a diagram for explaining the positional relationship between the area of the image captured by the camera 152 and the LEDs 151 a to 151 k of the light source unit 151. FIG. 12B is a diagram for explaining the relationship between the area on the image where the determination target object is located and the lighting control of the LEDs 151a to 151k when determining the three-dimensional object.

  As shown in FIG. 12A, areas divided by the reflection of the ribs 142a, 142b, 142c, 142d are area 1, area 2, area 3, area 4, and area 5 in order from the left of the image. say. Further, regarding the area 2, area 3 and area 4 excluding the areas at both ends, the inside of the area is further divided, and the left part is called L, the right part is called R, and the middle is called C.

  The LEDs 151a to 151k are arranged in consideration of the positions of the ribs 142a, 142b, 142c, and 142d. The LED 151c is directly below the rib 142a, the LED 151e is directly below the rib 142b, and the LED 151g and the rib 142d are directly below the rib 142c. An LED 151i is arranged directly below the LED 151i. In addition, 151 a is disposed near the left end of the light source unit 151, and 151 k is disposed near the right end of the light source unit 151. Further, an LED 151b is disposed near the center of area 1, LED 151d is disposed near the center of area 2, LED 151f is disposed near the center of area 3, LED 151h is disposed near the center of area 4, and LED 151j is disposed near the center of area 5.

  The three-dimensional object is determined by comparing two images captured by irradiating light from different directions to an area where the three-dimensional object may exist. This is performed by utilizing the fact that a difference appears in an image captured by influence. FIG. 12B shows an LED that is turned on when the first image is illuminated from the left side and an LED that is turned on when the second image is illuminated from the right side for each detection target area of the three-dimensional object. It is the table which prescribed | regulated.

  As shown in FIG. 12B, when the detection target area of the three-dimensional object is the area 1, the light source that is turned on when the first image is picked up is the LED 151a, and the light source that is turned on when the second image is picked up. Is the LED 151c. Further, when the detection target area of the three-dimensional object is L in area 2, the light sources that are turned on when the first image is picked up are LED 151c, LED 151d, and LED 151e, and the light sources that are turned on when the second image is picked up are LED 151e. Further, when the detection target region of the three-dimensional object is C in area 2, the light source that is turned on when the first image is captured is the LED 151c, and the light source that is turned on when the second image is captured is the LED 151e. Further, when the detection target area of the three-dimensional object is R in area 2, the light source that is turned on when the first image is captured is the LED 151c, and the light source that is turned on when the second image is captured is the LED 151c, the LED 151d, LED 151e.

  Hereinafter, in the case where area 3, area 4, and area 5 are the detection target areas of the three-dimensional object, the light source that is turned on when the first image is picked up and the light source that is turned on when the second image is picked up are also shown in FIG. As shown in b). In addition, although the number of LED to light is described in the table | surface of FIG.12 (b), although it is necessary to set beforehand also about each light quantity according to the irradiation direction to a detection target, here it is simplified. Therefore, it is omitted.

  Next, a detailed processing procedure of the three-dimensional object determination process (step S212) illustrated in FIG. 6 will be described. FIG. 13 is a flowchart showing a detailed processing procedure of the three-dimensional object determination process (step S212) shown in FIG.

  The three-dimensional object determination unit 155h determines which area shown in FIG. 12A belongs to the straight line detected by the edge detection unit 155e and determined to be appropriate by the straight line detection unit 155f (step S501). Further, the three-dimensional object determination unit 155h turns on the LED of the first image capturing light source corresponding to the area determined in step S501 in accordance with the definition of FIG. 12B (step S502). An image is captured (step S503). Further, the three-dimensional object determination unit 155h turns on the LED of the second image capturing light source corresponding to the area determined in step S501 in accordance with the definition of FIG. 12B (step S504), and the camera 152 performs the second operation. An image is captured (step S505).

  Next, the three-dimensional object determination unit 155h converts the collected first image and second image into a gray multi-tone image (step S506). Also, a difference image composed of pixel value differences for each pixel of the first image and the second image converted into a gray multi-tone image is generated (step S507), and the created difference image is converted into a predetermined pixel. Binarization is performed using the value as a threshold value (step S508).

  Next, the three-dimensional object determination unit 155h counts the number of pixels determined to have a difference in the binarized difference image created in step S508 (step S509). The three-dimensional object determination unit 155h determines that there is no three-dimensional object when the number of pixels counted in step S509 is smaller than a predetermined threshold, and when the number of pixels counted in step S509 is equal to or larger than the predetermined threshold. It is determined that there is a three-dimensional object (step S510), and the process ends.

  As described above, in the first embodiment, banknotes in a bundled state are received by the deposit / withdrawal port 140, images of the accepted banknotes in a bundled state are captured by the camera 152, and the ribs 142a, 142b, 142c, 142d of the captured images are captured. As a foreign object, a bound object added to a banknote bundled based on detection of a straight line appearing in a direction substantially parallel to the thickness direction of the banknote reflected in the captured image is targeted for the area excluding the area where the Since it is configured to detect, when a banknote is bundled by some bundle including metal and / or a material other than metal, the bundle is inserted into the deposit / withdrawal port 140 of the banknote depositing / dispensing machine 100. Can be detected as a foreign object.

  In the first embodiment, regarding the banding determination process, regarding the area between the two detected straight lines, if the area is banded, the area has a certain width, and the luminance change of the pixels in the area It has been explained that it is determined whether or not it is a bandage on the basis of the feature that the number of edge pixels in the thickness direction of the banknote bundle in the area is small. However, depending on the type of banknote and the state of the banknote, there is a possibility that an area that is not a bandage is erroneously detected as a bandage. Specifically, in a bill having a band-shaped white region at the end of the bill, when the bill end on the imaging surface side of the bill bundle is curled uniformly, depending on the positional relationship with the light source May be misrecognized as a bandage on the white area of the banknote. Therefore, in the second embodiment, an embodiment will be described in which the bandage can be correctly detected even in such a case.

  First, an outline of a more strict banding determination process according to the second embodiment will be described. FIG. 14 is a diagram for explaining the outline of the banding determination process according to the second embodiment.

  FIG. 14A shows an example of a banknote bundle having a white area at the left end of the banknote as indicated by a dotted square. Moreover, the deposit / withdrawal port 140 shown in Example 2 has the light source part 151 similarly to Example 1, and as shown to Fig.14 (a), the light source part 151 has the lower surface of a banknote bundle at the back of drawing. It is installed to irradiate from the side. The detailed positional relationship is shown in FIG. As shown in FIG. 14A, when a straight line is detected based on an image acquired from the y direction with respect to a bundle of banknotes having a single color area such as a band-like white color at the edge of the banknote. Even in the case where the banding shown in FIG. 14A does not exist, there is a possibility that the band-like white region is erroneously detected as a banding.

  FIG. 14B shows an example of an image obtained by capturing the banknote bundle shown in FIG. 14A from the y direction shown in FIG. Moreover, as shown in FIG.14 (b), the light source part 151 irradiates light with respect to a banknote bundle from the direction of the arrow which shows the bottom face of a banknote bundle on drawing. FIG.14 (c) is the figure which expanded the banding part shown with the dotted-line circle shown in FIG.14 (b). As shown in FIG. 14C, the bandage image has high brightness compared with other bandage areas near the corner of the bandage near the light source due to the relationship between the light source 151 and the bandage surface. It has the characteristics. In the second embodiment, the determination of the bundling is tightened based on the feature that if the bundling is performed as described above, the shine appears in the corner peripheral portion near the light source.

  FIG. 14D is a diagram showing a positional relationship between the light source unit 151 and an example of an image of a banknote bundle viewed from the x direction of the banknote bundle shown in FIG. In the case of an official seal, etc., as shown in FIG. 14 (d), there may be a case where a delicate curl in the same direction by cutting the banknote remains at the end of the banknote. When the light source 151 is positioned on the convex side of the curl, as shown in FIG. 14D, the curl direction and the light source 151 have a positional relationship between the white area and the pattern area on the banknote edge. May be detected as a straight line, or shine may be detected in the white area.

  FIG. 14E is an example when the number of banknotes inserted into the deposit / withdrawal port 140 is large, and an example of an image of a banknote bundle viewed from the x direction of the banknote bundle shown in FIG. It is the figure which showed the positional relationship with the light source part 151. FIG. As shown in FIG. 14 (e), when the number of banknotes is large, the amount of light for the area indicated by the arrow in FIG. 14 (e) is insufficient due to the positional relationship between the light source 151 and the banknote bundle with respect to the banknote bundle. Even if there is a bandage, there is a case where no shine occurs.

  In Example 2, in a banknote having a band-like white region in the short side direction of the banknote as shown in FIG. 14A, the shine is detected at a portion that is not a band seal as shown in FIG. 14D. By appropriately dealing with cases and cases that cannot be detected as shine even if it is a bandage as shown in FIG. 14 (e), based on the characteristics related to the shine caused by the bandage shown in FIG. 14 (c) , Make a stricter determination of the banding. As shown in FIG. 14D, a specific method for avoiding the detection of the shine in the non-banded portion and the specific method for detecting the shining in the case as shown in FIG. 14E will be described later. To do.

  Next, the internal structure and operation of the deposit / withdrawal port 140 shown in FIG. 2 according to the second embodiment will be described with reference to FIG. FIG. 15 is a diagram for explaining a coping method in the case where the state shown in FIG. 14E is reached when the number of banknotes inserted into the deposit / withdrawal port 140 is large.

  FIGS. 15A to 15E are views showing a cross section of the deposit / withdrawal port 140 by a plane perpendicular to the apparatus bottom face and the apparatus rear face of the bill depositing / dispensing machine 100. FIG. FIG. 15A shows a state in which a banknote bundle is inserted into the deposit / withdrawal port 140 and the shutter 141 is closed. Since the deposit / withdrawal port 140 is tilted forward, the inserted banknote bundle has a banknote surface as the front guide portion 145 of the deposit / withdrawal port 140 as viewed from the operator, and the rib 142 has a surface in the thickness direction of the banknote bundle. It is placed so that it touches. As shown to Fig.15 (a), since the light source part 151 and the camera 152 are arrange | positioned with respect to the mounted banknote bundle in the downward | lower side of the deposit or withdrawal port 140 seeing from an operator, When there is a bundling, a shine is detected around the corner of the banknote bundle indicated by the arrow. In addition, a feeding portion 144 is provided on the front side as viewed from the operator of the deposit / withdrawal port 140, and after confirming that foreign matters such as rubber bands and band seals are not included, the bills inserted into the deposit / withdrawal port 140 are inserted. One sheet is fed out to the transport unit 160.

  FIG. 15B is a view showing a state in which the banknote holder 143 moves in the direction indicated by the arrow and presses it toward the guide portion 145 side with respect to the banknote bundle inserted into the deposit / withdrawal port 140. In the state shown in FIG. 15B, the light source unit 151 irradiates light on the lower surface of the banknote bundle, and the camera 152 acquires an image of the lower surface of the banknote bundle. A link mechanism and a motor (not shown) are used for driving the bill presser 143.

  15C, after it is determined that there is no foreign object based on the image of the banknote bundle imaged in FIG. 15B, the guide unit 145 and the banknote holder 143 move together in the direction indicated by the arrow. FIG. 6 is a diagram illustrating a state in which a banknote is moved to a position where the banknote can be fed to the transport unit 160 by the feeding unit 144. Among the bills that are placed, the bill holder 143 rotates in the clockwise direction in FIG. 15C while the bill holder 143 presses the bill portion 143 toward the kicker roller 144a side of the feeding portion 144. The banknote on the lower left side in FIG. 15C is kicked downward one by one by the kicker roller 144a, and the kicked banknote is fed out one by one to the transport unit 160 by the feed roller 144b and the opposing roller 144c. become.

  FIG. 15D is a diagram showing an example when the number of inserted banknote bundles is large. When the image of the lower surface of the banknote bundle is acquired in the state shown in FIG. 15D, the image of the portion close to the corner of the banknote bundle indicated by the arrow is a dark image depending on the positional relationship between the light source unit 151 and the camera 152. As a result, there is a case where the shine that is the feature of the band seal cannot be detected. When the positional relationship between the banknote bundle, the light source unit 151, and the camera 152 is as shown in FIG. 15D, a straight line that is a candidate for banding is indicated by an arrow in FIG. 15D. In the case of being detected in the vicinity of the corner, the guide portion 145 and the bill presser 143 move integrally in the direction indicated by the arrow as shown in FIG. Then, the positional relationship between the corners of the banknote bundle indicated by the arrows, the light source unit 151 and the camera 152 is changed, the image of the lower surface of the banknote bundle is acquired again, and the presence / absence of the bandage is determined based on the acquired image. . By using the image acquired in this way, even if the number of inserted banknotes is large, if there is a bundling, it is possible to detect a shine on the bundling area, and to determine the bundling based on the detection of the bunch. It can be carried out.

  Next, an outline of a method for identifying the shining pixels of the banknote bundle according to the second embodiment will be described with reference to FIG.

  FIG. 16A is an example of an image of a portion corresponding to a band seal in the banknote bundle image captured by the camera 152. In the image shown in FIG. 16A, straight lines corresponding to the edge of the banding are detected at two places in the x-axis direction, and the y-axis on the side closer to the light source in the region sandwiched between the two straight lines. This is an example in which shine appears parallel to the x-axis near the end in the negative direction.

  A description will be given of a method for identifying a shine pixel that is a pixel constituting the shine for the image shown in FIG. For each linear pixel column parallel to the y-axis included in a region that is a candidate for banding of an image captured by the camera 152, whether or not each pixel included in the linear pixel column is a shine pixel By doing this, the shine pixel is specified.

  FIG. 16B is a diagram for explaining a method of specifying a shine pixel for each linear pixel column. In FIG. 16B, Y0, Y1, Y2, and Y3 indicate pixels included in the linear pixel row, and there are d1 pixels between the Y0 and Y1 pixels, and between the Y1 and Y2 pixels. Indicates that there are d2 pixels, and that there are d3 pixels between Y2 and Y3 pixels.

  The determination condition is, for example, that the luminance value of the Y3 pixel is higher than a predetermined threshold, the Y2 luminance value is higher than Y3, the Y1 luminance value is higher than Y2, the Y0 luminance value is lower than Y1, and the difference between the Y3 and Y1 luminance values. When the difference between the luminance values of Y1 and Y0 is larger than a predetermined threshold, Y1 is determined as a shine pixel.

  In this way, the shine pixels included in the linear pixel array parallel to the y-axis included in the region that is a candidate for banding are identified, and all the pixels parallel to the y-axis included in the region that is a candidate for banding are identified. By performing this process on all the linear pixel rows, it is possible to specify all the shining pixels included in the region that is a candidate for banding.

  Next, in the case where a banknote having a white area at the banknote edge shown in FIG. 14A is targeted, a mask range that is a non-evaluation area that does not search for a shining pixel used for banding determination is illustrated in FIG. 17 will be used for explanation. In the bundle of banknotes having a white area at the banknote edge shown in FIG. 14A, the banknote edge is curled as shown in FIG. 14D, and the convex surface of the curl faces the light source 151. In the case of a simple positional relationship or the like, there is a case where the shine is detected at a position corresponding to the white area at the edge of the bill even when the band seal is not present. In FIG. 17, assuming such a case, for the purpose of invalidating the shining pixel corresponding to the white area at the edge of the banknote in the banding determination, the shining pixel is not detected or the shining pixel is detected. A method for determining a mask range that is not used as invalid for the bandage determination will be described.

  When a bill having a white area as shown by a dotted square in FIG. 14 (a) is included in the bill end, it corresponds to the maximum width of the white region from the bill end considering all the types of bills to be considered. The maximum number of mask pixels, which is the number of pixels to be processed, can be stored in advance as known information.

  FIG. 17A is a diagram for explaining a method of determining a mask range when a straight line is detected within the range of the maximum number of mask pixels from the banknote end. First, the shine pixels in the area sandwiched between the detected straight line and the bill edge are specified. Next, the ratio of the shining pixels in the evaluation area surrounded by the detected straight line, the bill edge, and the range on the predetermined y axis, which is indicated by the dotted square in FIG. 17A, is calculated. When the ratio of the shine pixels calculated in this way is equal to or greater than a predetermined ratio, it is assumed that the identified shine is caused by the white area at the edge of the banknote, and the area between the detected straight line and the banknote edge is determined. As a mask range, the shining pixels in the range are excluded from the shining pixels used for the banding determination.

  FIG. 17B is a diagram for explaining the mask range when two straight lines are detected within the range of the maximum number of mask pixels from the banknote end. When two straight lines are detected within the range of the maximum number of mask pixels from the banknote edge, as shown by the dotted-line rectangle in FIG. The detected shine pixel is excluded from the shine pixels used for the band seal determination with the area between the straight line 1 detected at a position close to the bill edge and the bill edge as a mask range. As a result, when the detected straight line 1 and the straight line 2 detected at a position far from the bill edge are straight lines due to the banding, the shining pixels in the range between the straight line 1 and the straight line 2 are used for the banding determination. It will be used as a shining pixel to be used, and even if the bundling is present in the white region, the bundling using the shining can be determined.

  FIG. 17C shows that two straight lines are detected within the range of the maximum number of mask pixels from the bill edge, and the straight line 1 detected at a position close to the bill edge is effective only for a partial range in the thickness direction of the bill bundle. It is a figure for demonstrating the mask range in the case of having been determined. When it is determined that it is only effective halfway as in the detected straight line 1 shown in FIG. 17C, the y-coordinate range determined to be effective as a straight line is detected from the bill edge. With respect to the y-coordinate range determined to be not valid as a straight line with the range of the straight line 1 as the mask range, the area up to the straight line 2 detected at a position far from the bill edge is set as the mask range.

  Since the detection accuracy of the straight line detected due to the bundling is high, the possibility that a part of the straight line cannot be detected although it is a bundling is extremely low. In other words, even if the range of the y-coordinate determined to be not valid as a straight line is the mask range from the banknote edge to the x-coordinate corresponding to the straight line 1, the shine pixels in the range between the straight line 1 and the straight line 2 are found. It becomes effective, and the portion sandwiched between the straight line 1 and the straight line 2 is erroneously determined as a bandage. In order to prevent this, with respect to the range of the y-coordinate that has been determined not to be effective as a straight line, the area up to the straight line 2 detected at a position far from the bill edge is taken as the mask range.

  Next, an internal configuration of the deposit / withdrawal port foreign matter detection unit 350 of the banknote depositing / dispensing machine 100 according to the second embodiment will be described. The internal configuration of the deposit / withdrawal port foreign matter detection unit 350 of the banknote depositing / dispensing machine 100 according to the second embodiment is compared with the deposit / withdrawal port foreign matter detection unit 150 according to the first embodiment illustrated in FIG. The description will be made with reference to FIG. 5 by replacing the part 155e with the edge detection unit 355e and the band-seal determination unit 155g of Example 1 with the band-seal determination unit 355g. In the following, with respect to the edge detection unit 355e and the bundling determination unit 355g, which are different from the deposit / withdrawal port foreign matter detection unit 150 of the first embodiment, the difference from the edge detection unit 155e and the bundling determination unit 155g of the first embodiment will be emphasized. And explain.

  The edge detection unit 355e is the same as the first embodiment in the straight line detection method itself. The difference from the first embodiment is when there are more banknotes than a predetermined number as shown in FIG. The area close to the side indicated by the arrow in FIG. 15 (d) is around the optical axis of the attached light source, so that sufficient illuminance cannot be obtained and straight line detection is possible, but the accuracy of the determination of banding is insufficient. It becomes. When a straight line is detected in this region, the bill bundle is moved to the kicker roller 144a side by moving the guide portion 145 and the bill presser 143 integrally in the direction indicated by the arrow as shown in FIG. Move. Then, after again acquiring the image of the lower surface of the banknote bundle, straight line detection is performed again. It is determined that the number of banknotes is greater than a predetermined number by counting the number of pixels in the thickness direction of the banknote bundle based on the acquired image and determining that the number of pixels is equal to or greater than a predetermined threshold.

  The banding determination unit 155g according to the first embodiment determines that the area sandwiched between the two detected straight lines is banded with uniformity of the pixel value of the area sandwiched between the two detected straight lines. The band seal determination unit 355g of Example 2 detects a shine pixel in a predetermined range of a region sandwiched between the detected straight lines, and the ratio of the shine pixel to all the pixels in the predetermined range is equal to or greater than a predetermined threshold. The area sandwiched between the two detected straight lines is determined to be a bandage. However, in consideration of the case where the shining pixel is detected due to the influence of the white area from the bill edge, the shining pixel is not detected for the mask range described in FIG. Shall.

  Next, an overall processing procedure of foreign object detection processing by the deposit / withdrawal port foreign object detection unit 350 according to the second embodiment will be described. FIG. 18 is a flowchart illustrating the entire processing procedure of foreign object detection processing by the deposit / withdrawal port foreign object detection unit 350 according to the second embodiment. As described above, since the straight line detection method itself is the same as that of the first embodiment, the description of the portion described in the first embodiment is omitted by showing the correspondence with the processing described in the first embodiment. The description will focus on the differences from the first embodiment.

  The image capturing process (step S601) corresponds to the image capturing process (step S201) shown in the first embodiment. The straight line detection series process (step S602) corresponds to step S202 to step S208 of the first embodiment. Further, the branch to Yes in the straight line detection determination process (step S603) corresponds to the branch to Yes in step S206 in the first embodiment, and the branch to No in the straight line detection determination process (step S603) is performed. This corresponds to the branch to No in step S206 of Example 1. The continuity evaluation process (step S604) corresponds to step S209 in the first embodiment.

  Based on the thickness of the banknote bundle from the captured banknote bundle image, the banknote bundle is 200 or more, and it is determined to be effective in the evaluation of the continuity of the straight line in step S604 in the area near the light source 151 of the banknote bundle image. When there is a straight line (step S605; Yes), the edge detection unit 355e moves the guide unit 145 and the bill presser 143 integrally in the direction of the feeding unit 144 as shown in FIG. (Step S606), the process returns to Step S601. Moreover, even if it is determined that the banknote bundle is less than 200 sheets or the banknote bundle is determined to be 200 sheets or more, there is no straight line determined to be valid in step S604 on the side close to the light source unit 151 of the image of the banknote bundle. In the case (Step S605; No), if there is no straight line determined to be valid in the evaluation of the continuity of the straight line in Step S604 (Step S607; No), the processing is terminated assuming that no foreign matter is detected.

  If there is a straight line determined to be effective in the evaluation of the continuity of the straight line in step S604 (step S607; Yes), the banding determination unit 355g determines the maximum number of mask pixels, the detected straight line, and the validity of the straight line. Based on the range, a mask range in which no shining pixel detection is performed is determined (step S608). The details of the mask area determination processing are as described with reference to FIG. Further, the band seal determination unit 355g specifies a shine pixel for a region excluding the mask range determined in step S608 from the captured image (step S609). The bundling determination unit 355g determines the presence or absence of bundling based on the information on the straight line determined to be valid in the evaluation of the continuity of the straight line in step S604 and the shine pixel specified in step S609 (step S610). . Specifically, a shine pixel is detected in a predetermined range of a region sandwiched between two detected straight lines, and a ratio of the shine pixel to all pixels in the predetermined range is equal to or greater than a predetermined threshold value, It is determined that the region sandwiched between the two detected straight lines is a bandage.

  A three-dimensional object determination processing step S611 in step S611 corresponds to step S212 in the first embodiment.

  As described above, in the second embodiment, in addition to the configuration of the first embodiment, regarding the determination of the band seal, detection of the shining pixel is performed based on the predetermined range from the banknote edge in the longitudinal direction of the captured image and the detected straight line. A mask range to be excluded is specified, a shine pixel is detected in a region other than the mask range, and a ratio between the detected two straight lines is determined by a ratio of a predetermined range of shine pixels in the region between the two detected straight lines. Since it is determined that the region is a bandage, the detection of the bandage can be made stricter.

  In the first embodiment and the second embodiment described above, the light source unit 151 and the camera 152 are disposed below the deposit / withdrawal port 140, and the bill inserted into the deposit / withdrawal port 140 is imaged from below. However, the present invention is not limited to this, and a light source and a camera are attached to the upper part of the deposit / withdrawal port 140, and a mirror is disposed on the back surface of the shutter 141 to capture the bill reflected on the mirror surface. It is good also as what acquires the image from the top of a banknote. In addition, in order to detect a money clip of a non-metallic material sandwiched vertically, a light source and a camera are arranged beside the deposit / withdrawal port 140, and an image captured from the short direction of the inserted bill is acquired. Also good.

  Moreover, although the above-mentioned Example 1 and Example 2 demonstrated detecting the binding thing which binds this banknote bundle as a foreign material for the banknote which became the bundle state, this invention is limited to this. It is not a thing. For example, a bundle of valuable documents such as checks and gift certificates, and a piece of paper having a uniform size may be targeted, and a binding object that binds them may be detected as a foreign object.

  In the first and second embodiments described above, the captured image is converted into a gray image, and a straight line detection process, a continuity evaluation process, a banding determination process, and a three-dimensional object determination process are performed using the gray image. However, the present invention is not limited to this. For example, a saturation image or brightness image of a captured image is created, and a straight line detection process, a continuity evaluation process, a banding determination process, and a three-dimensional object determination process are performed using the saturation image or the brightness image. It is good. Further, an RGB image or a combination image thereof may be used. If the base image and band of the banknote in the captured image are about the same brightness, the gray image, lightness image, and saturation image will not have an edge, but if the two colors are different, the color image It is convenient to use Specifically, in the R image, the G image, and the B image, when the above-described straight line detection process, continuity evaluation process, banding determination process, and three-dimensional object determination process are performed, and a foreign object detection result is obtained in any image processing The result is the final determination result.

  Further, in the above-described first and second embodiments, the captured ribs 142a, 142b, 142c, and 142d are included in the areas including the portions where the ribs 142a, 142b, 142c, and 142d are captured, so that the captured ribs 142a are excluded. 142b, 142c, 142d are not erroneously detected as foreign matter added to the banknote, but the present invention is not limited to this. For example, when the material of the ribs 142a, 142b, 142c, and 142d is made of a material that transmits light, the reflection of the ribs 142a, 142b, 142c, and 142d does not occur, and the creation of the region that is not subject to foreign object detection is performed. The exclusion process for the reflection of the ribs 142a, 142b, 142c, and 142d may not be performed.

  In addition, each configuration illustrated in the above-described first and second embodiments is functionally schematic and does not necessarily need to be physically configured as illustrated. That is, the form of distribution / integration of each device is not limited to that shown in the figure, and all or a part thereof may be functionally or physically distributed / integrated in an arbitrary unit according to various loads or usage conditions. Can be configured.

  As described above, the foreign matter detection apparatus and foreign matter detection method according to the present invention are configured so that banknotes are deposited in a state in which banknotes are bound by some bundled material including metal and / or a material other than metal. It is suitable for detecting the bound material as a foreign object when it is put into the container.

DESCRIPTION OF SYMBOLS 100 Banknote deposit / withdrawal machine 110 Display operation part 120 Card reader 121 Card insertion slot 130 Number input part 140 Deposit / withdrawal slot 141 Shutter 142, 142a, 142b, 142c, 142d Rib 143 Bill holder 144 Conveyor 144a Kicker roller 144b Feed roller 144c Roller 145 Guide unit 150, 350 Deposit / withdrawal port foreign matter detection unit 151 Light source unit 151a, 151b, 151c, 151d, 151e, 151f, 151g,
151h, 151i, 151j, 151k LED
152 Camera 153, 210 Communication unit 154, 220 Storage unit 155, 230 Control unit 155a Automatic adjustment unit 155b Image imaging unit 155c Target area identification unit 155d Gray image generation unit 155e, 355e Edge detection unit 155f Straight line detection unit 155g, 355g Band seal determination Unit 155h three-dimensional object determination unit 160 transport unit 170 identification unit 180 bill storage unit 190 specification printer 191 specification issue port 200 voice communication unit 231 deposit processing unit 232 withdrawal processing unit

Claims (13)

A foreign matter detection device for detecting foreign matter thrown together with received paper sheets,
A paper sheet receiving means for receiving a bundle of paper sheets;
An image capturing means for capturing an image of the paper sheet in a bundle state from a direction in which the thickness of the bundle can be confirmed;
Target area specifying means for specifying a foreign substance detection target area from the image captured by the image capturing means;
Edge detection means for detecting an edge in a direction perpendicular to the thickness direction of the bundle of paper sheets from the image of the detection target area of the foreign matter specified by the target area specifying means;
Counting the number of pixels connected in a straight line of the edge image composed of edges detected by the edge detection means, and when there are more than a predetermined number of pixels, a straight line detection means for detecting a straight line from the edge image;
A foreign matter detection device comprising: a foreign matter detection means for detecting a bound object binding paper sheets as a foreign matter based on a straight line detected by the straight line detection means.   When there is an opaque member between the bundled paper sheets received by the paper sheet receiving means and the image capturing means, the target area specifying means is opaque from the image captured by the image capturing means. The foreign object detection device according to claim 1, wherein the region for the member is a non-target region.   The foreign matter detection apparatus according to claim 1, wherein the paper sheet receiving unit is transparent in a member existing between the received paper sheet in a bundled state and the image capturing unit.   4. The foreign object according to claim 1, wherein the image capturing unit captures an RGB color image, and detects the foreign object based on the captured R image, G image, and B image. Detection device.   The edge detection means generates a saturation image obtained by extracting saturation or a brightness image obtained by extracting lightness from an image captured by the image imaging means in a foreign object detection target area specified by the target area specifying means. The foreign object detection apparatus according to claim 1, wherein edge extraction is performed using the saturation image and / or the brightness image.   A light source that emits light in a single or a plurality of first directions and a light source that emits light in a second direction, each light source of a bundle of paper sheets received by the paper sheet receiving means Light irradiation control that is arranged at different positions with respect to the direction perpendicular to the thickness direction and controls light emission of a light source that irradiates light in the two directions onto the surface of a bundle of paper sheets to be imaged by the image imaging means The foreign object detection device according to claim 1, further comprising means.   The foreign object detection means is configured to turn on a light source that emits light in a first direction by the light irradiation control means, and to irradiate the paper sheets with light, and the first image captured by the image capturing means; The straight line detection is performed using the second image picked up by the image pickup means in a state in which a light source for irradiating light in the second direction is turned on by the light irradiation control means and the paper sheet is irradiated with light. The foreign object detection device according to claim 6, wherein a foreign object is detected based on an evaluation of a difference between the first image and the second image in an area including a straight line detected by the means. . The image capturing means can obtain a high resolution image and a low resolution image from the captured image,
The straight line detection means performs a straight line detection process based on the high resolution image acquired by the image imaging means when a straight line is detected based on the low resolution image acquired by the image imaging means. It performs. The foreign material detection apparatus as described in any one of Claims 1-7 characterized by the above-mentioned.   The foreign object detection means has a distance between two straight lines detected by the straight line detection means that is equal to or greater than a predetermined value, and a pixel value of an area sandwiched between the two straight lines of the image captured by the image capturing means is within a predetermined range. The foreign object detection device according to claim 1, wherein the foreign object is detected when it is contained. Thickness detecting means for detecting the thickness of the paper sheets in a bundled state received by the paper sheets receiving means;
A position changing means for changing the positional relationship between the light source and the paper sheet in a bundle state received by the paper sheet receiving means,
When the thickness of the sheet in the bundle state detected by the thickness detection unit is equal to or greater than a predetermined value and a straight line is detected by the straight line detection unit, the image imaging unit is bundled by the position changing unit. The foreign object detection according to claim 1, wherein the positional relationship between the paper sheet in a state and a light source is changed, and then an image of the paper sheet in a bundle state is re-captured. apparatus. A foreign matter detection method for detecting foreign matter thrown together with received paper sheets,
A paper sheet accepting step for accepting a bundle of paper sheets;
An image capturing step of capturing an image of the sheet in a bundle state from a direction in which the thickness of the bundle can be confirmed;
A target area specifying step for specifying a foreign substance detection target area from the image captured by the image capturing step;
An edge detection step for detecting an edge in a direction perpendicular to the thickness direction of the bundle of paper sheets from the image of the foreign object detection target area specified by the target area specifying step;
Counting the number of pixels connected in a straight line of the edge image composed of edges detected by the edge detection step, and when there are more than a predetermined number of pixels, a straight line detection step of detecting a straight line from the edge image;
A foreign matter detection method, comprising: a foreign matter detection step of detecting a bound object that binds paper sheets as a foreign matter based on the straight line detected by the straight line detection step. A foreign matter detection method for detecting a bandage wound around a loaded paper sheet as a foreign matter,
A paper sheet accepting step for accepting a bundle of paper sheets;
An image capturing step of capturing an image of the sheet in a bundle state from a direction in which the thickness of the bundle can be confirmed;
A target area specifying step for specifying a foreign substance detection target area from the image captured by the image capturing step;
An edge detection step for detecting an edge in a direction perpendicular to the thickness direction of the bundle of paper sheets from the image of the foreign object detection target area specified by the target area specifying step;
Counting the number of pixels connected in a straight line of the edge image composed of edges detected by the edge detection step, and when there are more than a predetermined number of pixels, a straight line detection step of detecting a straight line from the edge image;
In the straight line detecting step, when a straight line is detected within a predetermined distance from the edge of the paper sheet, the image of the paper sheet captured by the image capturing step is continuously from the edge of the paper sheet. If there is a region that maintains a uniform brightness with uniformity, a mask region determination that determines a region between the edge of the paper sheet and the straight line detected in the straight line detection step as a mask region Steps,
Detect pixels that are brighter than the pixels near the both ends in the thickness direction of the bundle of paper sheets in the areas other than the mask area determined in the mask area determination step, and that are brighter than the pixels in the thickness direction of the bundled paper sheet. Tekari pixel detection step,
A foreign matter detecting step comprising: detecting a straight line detected by the straight line detecting step; and a foreign matter detecting step of detecting, as a foreign matter, a bandage that binds paper sheets based on the shining pixel detected by the shining pixel detecting step. Detection method. The straight line detection step specifies a start point and an end point indicating a range in which the edge image and the effective edge image on which the straight line is based are present,
The mask area determination step includes
A mask area for each straight line, which is an area sandwiched between the edge of the paper sheet and the straight line, among the thickness direction areas sandwiched by the position in the thickness direction of the paper sheet corresponding to the start point and the end point with respect to the straight line Determine
When a plurality of the straight lines are detected within a predetermined distance from the edge of the paper sheet,
When the thickness direction regions of the straight-line mask regions corresponding to the respective straight lines overlap, the mask region corresponding to the overlapping thickness direction region is defined as the straight line of the straight line closest to the edge of the paper sheet. As an area corresponding to another mask area,
When the thickness direction regions of the straight line mask regions corresponding to the straight lines do not overlap, the mask regions for the non-overlapping thickness direction regions are overlapped with the straight line mask regions corresponding to all the straight lines. The foreign matter detection method according to claim 12, wherein the foreign matter detection method is a region.

Images