JP6936691B2 - Crack detection device, crack detection method, and computer program - Google Patents

Description

The techniques disclosed herein relate to crack detectors.

A technique for detecting cracks in an image of an object (for example, a wall of a building or a civil engineering structure) by performing image processing on an image of the object in order to diagnose the occurrence of cracks in the image. Is known (see, for example, Patent Document 1). According to the crack detection using such image processing, it is possible to efficiently and accurately diagnose the crack occurrence state in the object as compared with the case of visual inspection by a field worker, for example.

Japanese Unexamined Patent Publication No. 2012-98045

In the above-mentioned conventional technique, there is room for improvement in terms of crack detection accuracy. That is, in the above-mentioned conventional technique, a pixel that is not a pixel representing a crack but has a pixel value close to the pixel value of a pixel representing a crack (for example, a pixel representing a stain on the surface of an object) is erroneously detected as a crack. There are not a few false positives and omissions in which cracks that actually exist cannot be detected. Therefore, the above-mentioned conventional technique has a problem of suppressing the occurrence of such erroneous detection and detection omission and further improving the crack detection accuracy.

The present specification discloses a technique capable of solving the problem of improving the accuracy of crack detection using image processing.

The techniques disclosed herein can be realized, for example, in the following forms.

(1) The crack detection device disclosed in the present specification is a crack detection device that detects cracks in an object, and is composed of elements of N rows × N columns (where N is an odd number of 3 or more). The element value of each element overlapping the straight line passing through the central element located in the center of the row direction and the column direction is a positive value, and the element value of the element separated from the straight line in the direction orthogonal to the straight line by a predetermined distance or more. A line filter group composed of a plurality of line filters in which is 0 or less, and the slopes of the straight lines in each of the line filters are different from each other. The pixels constituting the image are sequentially selected as the pixels of interest, and for each of the line filters constituting the line filter group, the line filter is placed on the target image so that the central element of the line filter overlaps the pixels of interest. The original response value is calculated by performing a predetermined calculation using the element values of the elements of the line filter that are arranged and overlapped with each other and the pixel values of the pixels of the target image, and the pixel values of the attention pixels are set to each. A line filter processing unit that executes a line filter process that generates at least one post-processed target image by replacing it with a response value calculated based on the original response value calculated for the line filter, and the post-processed target image. A candidate determination unit is provided which sets a pixel whose pixel value is equal to or higher than a predetermined threshold value as a crack candidate pixel having a high probability of being a pixel representing a crack. In this crack detection device, the line filter processing unit executes line filter processing using the line filter group. Each line filter constituting the line filter group is composed of elements of N rows × N columns, and the element value of each element overlapping the straight line passing through the central element is a positive value, and the direction orthogonal to the straight line from the straight line. This is a filter in which the element value of an element separated by a predetermined distance or more is 0 or less. The slopes of the straight lines in which the elements whose element values are positive in each line filter are arranged are different from each other. In the line filter processing, the line filter processing unit sequentially selects the pixels constituting the target image as the pixels of interest, arranges the line filter on the target image for each line filter constituting the line filter group, and elements of the line filter. The original response value is calculated by executing a predetermined calculation using the element value of and the pixel value of the pixel of the target image, and the pixel value of the pixel of interest is set based on the original response value calculated for each line filter. By substituting with the calculated response value, at least one processed target image is generated. According to such a line filter process, a region in which pixels having a relatively large pixel value in the target image are lined up in a specific direction for a certain length or more (that is, a region having a line segment component having a certain length or more). In, the pixel value of the target image after the line filter processing becomes a relatively large value. Further, in the present crack detection device, the candidate determination unit determines, in the processed target image, pixels whose pixel values are equal to or greater than a predetermined threshold value, that is, pixels constituting a region having a line segment component having a certain length or longer. , Set as a crack candidate pixel. In general, a crack generated in an object contains a component of a line segment having a certain length or more (that is, a component that is elongated and extends linearly). Therefore, according to this crack detection device, a portion that is not a crack but has a pixel value close to the crack portion (for example, dirt on the surface of an object) is erroneously detected as a crack, or actually exists. It is possible to suppress the occurrence of detection omissions in which cracks cannot be detected, and it is possible to improve the accuracy of crack detection using image processing.

(2) In the crack detection device, the predetermined calculation calculates a power value having the pixel value as the base and the element value as a power index, and the power value of all the elements constituting the line filter. The operation may be such that the product is the original response value. According to this crack detection device, in a region where pixels having a relatively large pixel value in a target image are lined up in a specific direction for a certain length or more (that is, a region having a line segment component having a certain length or more). The pixel value of the target image after the line-filter processing becomes a large value that can be more clearly distinguished from the pixel value of the other region. Therefore, according to this crack detection device, it is possible to effectively suppress the occurrence of erroneous detection and detection omission, and it is possible to effectively improve the accuracy of crack detection using image processing.

(3) In the crack detection device, the predetermined calculation calculates a multiplication value of the pixel value and the element value, and the sum of the multiplication values for all the elements constituting the line filter is the original response. It may be configured as a value operation. According to this crack detection device, in a region where pixels having a relatively large pixel value in a target image are lined up in a specific direction for a certain length or more (that is, a region having a line segment component having a certain length or more). The pixel value of the target image after the line-filter processing becomes a large value that can be more clearly distinguished from the pixel value of the other region. Therefore, according to this crack detection device, it is possible to effectively suppress the occurrence of erroneous detection and detection omission, and it is possible to effectively improve the accuracy of crack detection using image processing.

(4) In the crack detection device, the line filter processing unit generates the processed target image for each line filter, and is calculated for generating the processed target image for one of the line filters. The response value is a value obtained by multiplying the original response value calculated by using the one line filter by a first weighting coefficient, and at least one whose slope of the straight line is close to that of the one line filter. The configuration may be the sum of the original response values calculated by using the other line filter and each value obtained by multiplying the original response value by a weighting coefficient smaller than that of the first weighting coefficient. According to this crack detection device, cracks can be detected accurately even when the direction of cracks in the target image deviates slightly from the inclination of the straight line in the line filter, and the accuracy of crack detection using image processing can be improved. It can be effectively improved.

(5) In the crack detection device, the response value may be the maximum value among the original response values calculated for each line filter. According to this crack detection device, cracks in a target image can be detected with high accuracy, and the accuracy of crack detection using image processing can be effectively improved.

(6) In the crack detection device, in the line filter, the element value of each element overlapping the straight line is the first positive value, and the line filter is separated from the straight line by a first distance or more in the direction orthogonal to the straight line. A second positive value in which the element value of the element is 0 or less and the element value of the element separated from the straight line by less than the first distance in the direction orthogonal to the straight line is smaller than the first positive value. It may be configured as a filter which is. According to this crack detection device, cracks having a certain width in the target image can be detected with high accuracy, and the accuracy of crack detection using image processing can be effectively improved.

(7) In the crack detection device, the storage unit stores a plurality of the line filter groups having different N values from each other, and the line filter processing unit selects one of the plurality of line filter groups. And, if the target image is targeted, or if the line filter processing is executed at least once, the target image after the processing after the line filtering processing executed immediately before is selected. The line filter process may be executed repeatedly using the line filter group. According to this crack detection device, it is possible to effectively suppress the occurrence of erroneous detection and detection omission as compared with the case where the line filter processing is executed only once, and the crack detection using image processing is performed. The accuracy of can be effectively improved.

(8) In the crack detection device, the line filter processing unit may be configured to select the line filter group in descending order of the value of N. According to this crack detection device, it is possible to more reliably suppress the occurrence of erroneous detection in which noise pixels are erroneously detected as crack pixels, and it is possible to effectively improve the accuracy of crack detection using image processing. ..

(9) The crack detection device may further include a display control unit that displays an image showing the crack candidate pixels on the display. According to this crack detection device, the user can grasp the occurrence status of cracks in the object.

(10) In the crack detection device, the object may be a wall-shaped structure. According to this crack detection device, crack detection is performed on a wall-shaped structure that is an object in which crack generation is likely to be a problem and in which crack detection needs to be performed in a relatively wide range. Can be executed efficiently and accurately.

The techniques disclosed in the present specification can be realized in various forms, for example, a crack detection device, a crack detection method, a computer program for realizing the functions of those devices or methods, and the like. It can be realized in the form of a non-temporary recording medium or the like on which a computer program is recorded. Further, the technique disclosed in the present specification is also useful as an elemental technique of an automatic inspection system using a robot.

It is explanatory drawing which shows schematic structure of the crack detection apparatus 100 in 1st Embodiment. It is explanatory drawing which shows an example of the screen (processing screen S1) which is displayed on the display unit 120 at the time of execution of a crack detection process. It is a flowchart which shows the flow of the crack detection process. It is explanatory drawing which shows an example of the processing screen S1 including the result of the crack detection processing. It is a flowchart which shows the flow of the crack candidate detection processing. It is explanatory drawing which shows an example of a line filter group LFG. It is explanatory drawing which shows an example of the target image IO. It is a flowchart which shows the line filter processing using the line filter group LFG. It is explanatory drawing which shows the outline of the line filter processing using the line filter group LFG. It is explanatory drawing which shows an example of the setting result of the crack candidate region CC. It is a flowchart which shows the flow of the crack candidate confirmation processing. It is explanatory drawing which shows the outline of the crack candidate confirmation processing. It is a flowchart which shows the flow of the crack candidate detection processing in 2nd Embodiment. It is explanatory drawing which shows an example of the line filter group LFG in 2nd Embodiment. It is a flowchart which shows the line filter processing using the line filter group LFG in 2nd Embodiment. It is explanatory drawing which shows the outline of the line filter processing using the line filter group LFG in 2nd Embodiment.

A. First Embodiment:
A-1. Device configuration:
FIG. 1 is an explanatory diagram schematically showing the configuration of the crack detection device 100 according to the first embodiment. The crack detection device 100 is a general-purpose computer such as a personal computer, a tablet terminal, or a smartphone. The crack detection device 100 includes a storage unit 110, a display unit (display) 120, an input unit 130, an interface unit 140, and a control unit 170. Each of these parts is communicably connected to each other via a bus 190.

The display unit 120 is composed of, for example, a liquid crystal display or the like, and displays various images and information. The input unit 130 is composed of, for example, a keyboard, a mouse, a microphone, or the like, and receives user operations and voice instructions. The interface unit 140 is composed of, for example, a LAN interface, a USB interface, or the like, and communicates with another device (for example, an image pickup device 320) by wire or wirelessly.

The storage unit 110 is composed of, for example, a hard disk drive (HDD) or the like, and stores various programs and data. For example, the storage unit 110 stores a crack detection program CP for executing a crack detection process described later. The crack detection program CP is provided in a state of being stored in a computer-readable recording medium (not shown) such as a CD-ROM, a DVD-ROM, or a USB memory, and is installed in the crack detection device 100 to store the crack detection program CP. It is stored in 110.

Further, the image data Id is stored in the storage unit 110. The image data Id is generated by, for example, capturing an object (in this embodiment, the wall of the building 310) with an image pickup device 320 such as a digital still camera. The image data Id is input to the crack detection device 100 via the interface unit 140 and stored in the storage unit 110.

The control unit 170 is composed of, for example, a CPU, a ROM, a RAM, or the like, and controls the operation of the crack detection device 100 by executing a computer program read from the storage unit 110. For example, the control unit 170 functions as a crack detection processing unit 200 that executes a crack detection process described later by reading and executing the crack detection program CP. The crack detection processing unit 200 includes a display control unit 210, a candidate detection processing unit 220, and a candidate confirmation processing unit 230. Further, the candidate detection processing unit 220 includes a line filter processing unit 222, a candidate determination unit 224, and a candidate area setting unit 226. Further, the candidate confirmation processing unit 230 includes a rectangle calculation unit 232 and a confirmation determination unit 234. The functions of each of these parts will be described in accordance with the description of the crack detection process described later.

A-2. Crack detection process:
Next, the crack detection process executed by the crack detection device 100 will be described. The crack detection process is a process of detecting cracks in an image obtained by capturing an image of an object by performing image processing on the image. In general, a crack generated in an object contains a component of a line segment (that is, a component that is elongated and extends linearly). In the present embodiment, the crack detection process with high accuracy is realized by utilizing such a feature of the crack. In the present embodiment, the object of the crack detection process is a tiled wall (that is, a wall-shaped structure) of the building 310.

FIG. 2 is an explanatory diagram showing an example of a screen (hereinafter, referred to as “processing screen S1”) displayed on the display unit 120 when the crack detection process is executed. The processing screen S1 is displayed on the display unit 120 by the display control unit 210. The processing screen S1 includes an image selection area R1, a method selection area R2, and an image display area R3. The image selection area R1 is an area for selecting a target image IO to be subjected to the crack detection process. The method selection area R2 is an area for selecting a method (algorithm) used for the crack detection process. The image display area R3 is an area for displaying the target image IO and the like.

In the image selection area R1 of the processing screen S1, an image identifier (file name, thumbnail, etc.) represented by the image data Id (see FIG. 1) stored in the storage unit 110 is displayed as an option. When the user selects a desired image (for example, "image 1") from the image choices displayed in the image selection area R1 via the input unit 130, the selected image is set as the target image IO. It is displayed in the image display area R3. The image selection area R1 includes a number display column R11 for displaying the number of cracks detected by the crack detection process.

Further, in the method selection area R2 of the processing screen S1, options of the method (algorithm) used for each of the processes (crack candidate detection process, crack candidate confirmation process) described later that constitute the crack detection process are displayed. From the plurality of methods displayed in the method selection area R2, the user uses the input unit 130 to perform a desired method for each process (for example, "method A" of the crack candidate detection process and the crack candidate confirmation process. When "Method H") is selected and the "Execute" button B1 is selected, the crack detection process using the selected method is started. As shown in FIG. 2, in the present embodiment, the target image IO is an image showing the tiled wall of the building 310, and each pixel constituting the image is represented by a pixel value of 256 gradations. This is a grayscale image. In this grayscale image, the pixel value of the black pixel is the minimum value (that is, 0), and the pixel value of the white pixel is the maximum value (that is, 255). Further, the target image IO includes a pixel representing a crack (hereinafter referred to as "crack pixel Pc") and a pixel that is not a crack pixel Pc but has a pixel value close to the pixel value of the crack pixel Pc (for example, an object). It is a pixel representing dirt on the surface and the like, and includes (hereinafter, referred to as “noise pixel Pn”).

FIG. 3 is a flowchart showing the flow of the crack detection process. First, the crack detection processing unit 200 executes preprocessing for the target image IO, if necessary (S120). As preprocessing, for example, a process of correcting the pixel value of the target image IO, an image area representing a tile (hereinafter referred to as "tile area") is extracted from the target image IO, and an image area (joint area) other than the tile area is extracted. Areas, etc.) may be excluded from the target range for crack detection. The process of extracting the tile region can be executed by using a known method. Further, when the target range for crack detection in the target image IO is predetermined, it is not necessary to execute the process of extracting the tile area.

Next, the candidate detection processing unit 220 executes the crack candidate detection processing (S130). The crack candidate detection process is a process of detecting a crack candidate region CC including a crack candidate pixel Pcc, which is a pixel representing a crack, from the target image IO. The details of the crack candidate detection process will be described later.

Next, the crack detection processing unit 200 determines whether or not at least one crack candidate region CC is detected in the crack candidate detection process (S130) (S140). When it is determined that the crack candidate region CC has been detected (S140: YES), the candidate confirmation processing unit 230 executes the crack candidate confirmation process (S150). The crack candidate confirmation process is a process for confirming whether or not the crack candidate region CC detected from the target image IO is the crack region CR, which is an image region that truly represents a crack. The details of the crack candidate confirmation process will be described later. If it is determined that the crack candidate region CC is not detected in the crack candidate detection process (S140: NO), the crack candidate confirmation process (S150) is skipped.

Next, the display control unit 210 causes the display unit 120 to display the result of the crack detection process (S160). FIG. 4 is an explanatory diagram showing an example of the processing screen S1 including the result of the crack detection processing. In the example of the processing screen S1 shown in FIG. 4, an image showing the detected crack region CR on the target image IO displayed in the image display region R3 (that is, an image showing at least a part of the crack candidate pixels Pcc). Is displayed. Further, in the number display field R11 of the image selection area R1 of the processing screen S1, the number of cracked areas CR detected from the target image IO (for example, one) is shown. The user can grasp the occurrence state of cracks in the object (for example, the wall of the building 310) represented by the target image IO through the processing screen S1 showing the result of the crack detection process.

A-3. Crack candidate detection process:
Next, the crack candidate detection process (S130 in FIG. 3) will be described in detail. As described above, the crack candidate detection process is a process of detecting a crack candidate region CC including a crack candidate pixel Pcc, which is a pixel representing a crack, from the target image IO. FIG. 5 is a flowchart showing the flow of the crack candidate detection process.

As shown in FIG. 1, the line filter group LFG is stored in the storage unit 110 of the crack detection device 100. In the present embodiment, the storage unit 110 stores a plurality of line filter groups LFG having different numbers of elements on one side (hereinafter, referred to as “number of unit elements”) of the line filter LF, which will be described later. In the following description, the line filter group LFG having a unit number of elements N is referred to as a line filter group LFG (N). The number of unit elements N is an odd number of 3 or more.

FIG. 6 is an explanatory diagram showing an example of the line filter group LFG. FIG. 6 illustrates a line filter group LFG (7) in which the number of unit elements N is 7, and an LFG (9) in which the number of unit elements N is 9. The number of line filter groups LFG stored in the storage unit 110 and the number N of unit elements of each line filter group LFG can be arbitrarily set.

The line filter group LFG (N) having the number of unit elements N is composed of a plurality of line filters LF (N) composed of elements of N rows × N columns. In each line filter LF (N), the element value of each element overlapping the straight line passing through the central element CE is a positive value, and the element value of the element separated from the straight line in the direction orthogonal to the straight line by a predetermined distance or more. Is a filter in which is 0 or less. In the present embodiment, in each line filter LF (N), the element value of each element overlapping the straight line passing through the central element CE (each element constituting the element string L0 in FIG. 6) is 1, and two elements from the straight line. The element value of each separated element (each element constituting the element sequence L2 in FIG. 6) is 0, and each element separated by three elements from the straight line (each element constituting the element sequence L3 in FIG. 6) The element value is −0.2, and the element value of each element (each element constituting the element sequence LX in FIG. 6) separated from the straight line by 4 elements or more is 0. Further, in the present embodiment, in each line filter LF (N), the element value of each element separated by one element from the straight line passing through the central element CE (each element constituting the element sequence L1 in FIG. 6) is 0.2. Is. In each line filter LF (N) used in the present embodiment, the distance for two elements corresponds to the first distance in the claims, and the element value "1" is the first in the claims. It corresponds to a positive value, and the element value "0.2" corresponds to a second positive value in the claims.

Further, each of the plurality of line filters LF (N) constituting a certain line filter group LFG (N) has different slopes of the straight lines (hereinafter, referred to as “filter angles”) in which the elements having the element value of 1 are arranged. ing. In the present embodiment, the line filter group LFG (N) is composed of 12 line filters LF (N) having filter angles different from each other by 30 degrees. For example, as shown in FIG. 6, in one line filter LF (7) -0 constituting the line filter group LFG (7) in which the number of unit elements N is 7, the element having an element value of 1 is in the horizontal direction ( The other line filter LF (7) -90 is arranged in the horizontal direction (that is, the filter angle is 0 degrees), and the element having the element value of 1 is in the vertical direction (vertical direction in the figure). ) (Ie, the filter angle is 90 degrees). The line filter group LFG (7) also includes a line filter LF (7) -30 having a filter angle of 30 degrees, a line filter LF (7) -60 having a filter angle of 60 degrees, and the like. .. The same applies to other line filter groups LFG.

When the crack candidate detection process (FIG. 5) is started, the line filter processing unit 222 of the candidate detection processing unit 220 performs a black-and-white inversion process on the target image IO. As a result, the target image IO becomes an image in which the pixel value of the black pixel is the maximum value (that is, 255) and the pixel value of the white pixel is the minimum value (that is, 0). Next, the line filter processing unit 222 of the candidate detection processing unit 220 selects one line filter group LFG from the plurality of line filter group LFGs stored in the storage unit 110 (S210). In the present embodiment, the line filter processing unit 222 selects the line filter group LFG having the largest number of unit elements N among the unselected line filter group LFGs. The line filter processing unit 222 acquires the selected line filter group LFG by reading it from the storage unit 110.

Next, the line filter processing unit 222 executes line filter processing using the selected line filter group LFG for the target image IO (S220). FIG. 7 is an explanatory diagram showing an example of the target image IO. FIG. 7 shows an enlarged configuration of a part of the target image IO (a part corresponding to the X1 part of FIG. 2). In the example shown in FIG. 7, the target image IO includes a crack pixel Pc representing a crack and a background pixel Pb representing a background (normal tile portion) other than the crack. The cracked pixel Pc has a relatively large pixel value (that is, is relatively close to the black pixel value), and the background pixel Pb has a relatively small pixel value (that is, is relatively close to the white pixel value). Further, in the example shown in FIG. 7, the target image IO is a pixel that is not the cracked pixel Pc but has a pixel value close to the pixel value of the cracked pixel Pc (for example, a pixel representing dirt on the surface of the object). The noise pixel Pn is also included.

FIG. 8 is a flowchart showing a line filter process using the line filter group LFG. Further, FIG. 9 is an explanatory diagram showing an outline of line filter processing using the line filter group LFG. In column A of FIG. 9, the pixel value of each pixel constituting (a part of) the target image IO is shown. In this example, the pixel value is relatively large in a region (area surrounded by a broken line) extending in the vertical direction (vertical direction in the figure) located near the center of the target image IO. This region is a crack region CR composed of crack pixels Pc. Further, in column B of FIG. 9, an example of a filter calculation using the line filter LF (7) -90 (see FIG. 6) in which the number of unit elements N is 7 and the filter angle is 90 degrees is shown. .. Further, in column C of FIG. 9, the target image IO after the filter calculation using the line filter LF (7) -90 is shown. The target image IO after the filter calculation using the line filter having the filter angle θ (degrees) is referred to as the target image IO (f) −θ after the filter calculation.

In the line filter processing (FIG. 8) using the line filter group LFG, first, the line filter processing unit 222 selects one of the pixels constituting the target image IO as the pixel of interest SP (S221). In column A of FIG. 9, a state in which one of the pixels constituting the target image IO is selected as the pixel of interest SP is shown.

Next, the line filter processing unit 222 selects one of the line filter LFs constituting the selected line filter group LFG (S222), and arranges the selected line filter LF on the target image IO (S223). .. At this time, the central element CE of the line filter LF is overlapped with the pixel of interest SP. In column A of FIG. 9, the positions of the line filters LF (7) -90 arranged on the target image IO are indicated by thick solid lines.

Next, the line filter processing unit 222 calculates the original response value Ro by executing a predetermined filter calculation using the element values of the elements of the line filter LF that overlap each other and the pixel values of the pixels of the target image IO. (S224). Specifically, the line filter processing unit 222 calculates a power value that should have the pixel value of the pixel of the target image IO as the base and the element value of the line filter LF as a power index, and for all the elements constituting the line filter LF. The product of the above exponentiation values is calculated as the original response value Ro. For example, in the first step of S222 in the line filter processing using the line filter group LFG (7) in which the number of unit elements N is 7, the line filter LF (7) −90 having a filter angle of 90 degrees shown in FIG. Is selected. In this case, the original response value Ro is 30 ^−0.2 × 300 ⁰ × 150 ^0.2 × 200 ¹ × ・ ・ ・ × 30 ^−0.2 × 30 ^{−, as shown in column B of FIG. 0.2} × 300 ⁰ × 150 ^0.2 × 200 ¹ × ・ ・ ・ × 30 ^−0.2 = 1.2 × 10 ¹⁸ .

After calculating the original response value Ro (S224), the line filter processing unit 222 determines whether or not all the line filter LFs constituting the currently selected line filter group LFG have been selected (S225). If it is determined that there is an unselected line filter LF (S225: NO), the process returns to S222 and the above processing is repeated. For example, in the second step of S222 in the line filter processing using the line filter group LFG (7) in which the number of unit elements N is 7, the line filter LF (7) -0 having a filter angle of 0 degrees shown in FIG. Is selected. In this case, the line filter LF (7) -0 is arranged on the target image IO (S223), and the original response value Ro is calculated (S224). At this time, the original response value Ro ^{is, 30 -0.2 × 30 -0.2 × 150} -0.2 × 200 -0.2 × ··· × 30 -0.2 × 30 0 × 30 0 × 150 ⁰ × 2000 ⁰ × ・ ・ ・ × 30 ^−0.2 = 3.6 × 10 ¹² .

Here, the fact that the original response value Ro calculated for the line filter LF of a certain filter angle is large means that the direction (element) of the filter angle of the line filter LF in the image region overlapping the line filter LF in the target image IO. It is the direction of a straight line in which elements having a value of 1 are lined up, and means that pixels having a relatively large pixel value (that is, close to black) are lined up in a line filter LF (7) -90, for example, in the vertical direction). .. For example, in the example shown in column A of FIG. 9, pixels having relatively large pixel values are arranged in the vertical direction in the image region overlapping the line filter LF in the target image IO. ^{Therefore, in this example, the original response value Ro (1.2 × 10 18} ) calculated by using the line filter LF (7) −90 having a filter angle of 90 degrees is a line filter having a filter angle of 0 degrees. It is larger than the original response value Ro (3.6 × 10 ^{12) calculated using LF (7) -0.} As described above, in the line filter processing of the present embodiment, in the target image IO, a pixel having a relatively large pixel value centered on the pixel of interest SP has a predetermined length (a length corresponding to the number of unit elements N) in a specific direction. ) When they are lined up as above, the original response value Ro for the line filter LF having a filter angle corresponding to the direction becomes a relatively large value.

The line filter processing unit 222 repeatedly executes the processes from the selection of the line filter LF (S222) to the calculation of the original response value Ro (S224), and all the line filter LFs constituting the currently selected line filter group LFG. (That is, for the currently selected pixel SP of interest, the calculation of the original response value Ro for all the line filters LF has been completed) (S225: YES), the target image IO It is determined whether or not the selection as the pixel of interest SP is completed for all the pixels constituting the above (S227), and if it is determined that there are unselected pixels (S227: NO), the process returns to S221 and described above. Repeat the process. That is, one of the unselected pixels in the target image IO is newly selected as the pixel of interest SP, and the newly selected pixel of interest SP is subjected to line filter processing using the currently selected line filter group LFG. By doing so, the original response value Ro for each line filter LF is calculated.

By repeating such processing, the original response value Ro for each line filter LF constituting the currently selected line filter group LFG is calculated for all the pixels constituting the target image IO. As described above, after the filter calculation, the pixel value of each pixel constituting the target image IO is replaced with the original response value Ro calculated by the filter calculation using the line filter of the filter angle θ (degrees). The target image is called IO (f) -θ. In column C of FIG. 9, a part of the target image IO (f) -90 after the filter calculation using the line filter LF (7) -90 is shown. In the target image IO (f) -90 after this filter calculation, the original response value Ro of the region (the region corresponding to the crack region CR) extending in the vertical direction (vertical direction in the figure) located near the center becomes a relatively large value. It has become.

When the line filter processing unit 222 determines that the selection as the pixel of interest SP has been completed for all the pixels constituting the target image IO (S227: YES), for each line filter LF (for each filter angle), By calculating the weighted response value Rw based on the original response value Ro and replacing the pixel value of each pixel of the target image IO with the weighted response value Rw, the processed target image for each line filter LF (for each filter angle) is generated. (S228). After that, the line filter processing is finished. The weighted response value Rw corresponds to the response value in the claims.

The weighted response value Rw is calculated as follows. For example, the weighted response value Rw for a filter angle of 90 degrees is the value obtained by multiplying the original response value Ro calculated using the line filter LF with a filter angle of 90 degrees by the first weighting coefficient W1 and the filter angle of 90 degrees. A second weighting factor W2 (provided that the second weighting factor W2 < It is the sum of the value obtained by multiplying the first weighting factor W1). The first weighting factor W1 is set to, for example, "1", and the second weighting factor W2 is set to, for example, "0.2". The number of other filter angles used in calculating the weighted response value Rw for a certain filter angle is not limited to two, and may be one or three or more. For example, when calculating the weighted response value Rw for a filter angle of 90 degrees, in addition to the original response value Ro calculated using the line filter LF with filter angles of 60 degrees and 120 degrees, the filter angle is 30 degrees and The original response value Ro calculated using the line filter LF of 150 degrees may be used. At that time, the value of the second weighting coefficient W2 to be multiplied by the original response value Ro calculated by using the line filter LF having the filter angles of 30 degrees and 150 degrees has the filter angles of 60 degrees and 120 degrees. It may be the same as the value of the second weighting coefficient W2 multiplied by the original response value Ro calculated by using the line filter LF of the above, or may be smaller than the value of the second weighting coefficient W2.

As described above, in the line filter processing (FIG. 8) of the present embodiment, in the target image IO, pixels having a relatively large pixel value centered on the sequentially selected pixel of interest SP have a predetermined length (unit) in a specific direction. When there are locations that are lined up at least (the length corresponding to the number of elements N), the pixel value of the pixel of interest SP is determined by the weighted response value Rw calculated for the line filter LF (filter angle) corresponding to the specific direction. It is replaced with a relatively large value, and when there is no such portion, the pixel value of the pixel of interest SP is replaced with a relatively small value. Therefore, in the line filter processing of the present embodiment, after the line filter processing, in a region (a region having a line segment component) in which pixels having a relatively large pixel value in the target image IO are lined up in a specific direction for a certain length or more. The pixel value of the target image IO for the line filter LF corresponding to the specific direction is relatively large.

As shown in FIG. 5, when the line filter processing (S220) using the selected line filter group LFG is completed, the candidate determination unit 224 determines the pixel value in the target image IO after the line filter processing in advance. A pixel having a threshold value or more is set as a crack candidate pixel Pcc having a high probability of being a pixel representing a crack (S240). As described above, in the line filter processing of the present embodiment, a region in which pixels having a relatively large pixel value in the target image IO are lined up in a specific direction for a certain length or more (that is, a line segment having a certain length or more). In the region having a component), since the pixel value of the target image IO after the line filter processing becomes a relatively large value, the pixel whose pixel value is equal to or larger than the threshold value is set as the crack candidate pixel Pcc.

Next, the candidate area setting unit 226 sets the image area including the set crack candidate pixel Pcc as the crack candidate area CC having a high probability of being the image area representing the crack (S250). The crack candidate region CC may be composed of only the crack candidate pixels Pcc, or may include a small number of pixels that are not the crack candidate pixels Pcc. For example, when there are a few pixels that are not the crack candidate pixels Pcc inside the region surrounded by the group of crack candidate pixels Pcc, the crack candidate region CC including the pixels that are not the crack candidate pixels Pcc is included. It may be set.

FIG. 10 is an explanatory diagram showing an example of the setting result of the crack candidate region CC. In FIG. 10, hatching is attached to the pixel (crack candidate region CC) set in the crack candidate pixel Pcc. In the example shown in FIG. 10, most of the crack pixel Pc (shown by the broken line) that truly represents the crack is set to the crack candidate pixel Pcc. This is because the image region composed of the cracked pixels Pc contains components of relatively long line segments. On the other hand, all the noise pixels Pn (shown by the broken line) are not set in the crack candidate pixel Pcc. That is, the occurrence of erroneous detection in which these noise pixels Pn are erroneously detected as cracks is prevented. The image region composed of the noise pixels Pn does not include a component of a relatively long line segment as included in the image region composed of the crack pixels Pc, and therefore is not set as the crack candidate pixel Pcc.

Next, the line filter processing unit 222 determines whether or not the selection of all the line filter groups LFG stored in the storage unit 110 is completed (S260), and determines that there is an unselected line filter group LFG. In the case (S260: NO), the process returns to S210 and the above processing is repeated. As described above, when the line filter group LFG is selected (S210), the line filter group LFG having the largest number of unit elements N is selected from the unselected line filter group LFGs. Further, when the crack candidate region CC is set in the above-mentioned processing using each of the plurality of line filter groups LFG, the final crack candidate region CC is calculated by adding up the crack candidate region CCs. Set.

By the crack candidate detection process described above, the crack candidate region CC including the crack candidate pixel Pcc, which is a pixel representing a crack and has a high probability, is detected from the target image IO.

A-4. Crack candidate confirmation process:
Next, the crack candidate confirmation process (S150 in FIG. 3) will be described in detail. As described above, the crack candidate confirmation process is a process for confirming whether or not the crack candidate region CC detected from the target image IO is the crack region CR, which is an image region that truly represents a crack. FIG. 11 is a flowchart showing the flow of the crack candidate confirmation process. Further, FIG. 12 is an explanatory diagram showing an outline of the crack candidate confirmation process.

First, the candidate confirmation processing unit 230 selects one crack candidate region CC (S310). FIG. 12 shows one selected crack candidate region CC. As described above, in the present embodiment, the crack candidate region CC is an image region composed of the crack candidate pixels Pcc.

Next, the rectangle calculation unit 232 of the candidate confirmation processing unit 230 calculates the minimum rectangle RE including the selected crack candidate region CC (S320). FIG. 12 shows the smallest rectangle RE that includes the selected crack candidate region CC.

Next, in the confirmation determination unit 234 of the candidate confirmation processing unit 230, the ratio of the length Lr (however, Lr ≧ Wr) to the width Wr of the minimum rectangle RE (hereinafter, referred to as “aspect ratio (Lr / Wr)”) is determined. It is determined whether or not it is equal to or higher than a predetermined threshold value Th1 (S330). The width Wr of the minimum rectangle RE corresponds to the width of the crack candidate region CC, and the length Lr of the minimum rectangle RE corresponds to the length of the crack candidate region CC.

When the aspect ratio (Lr / Wr) of the minimum rectangle RE is equal to or greater than the threshold Th1 (S330: YES), the confirmation determination unit 234 determines that the selected crack candidate region CC is a crack region CR that truly represents a crack. It is determined that there is (S340). On the other hand, when the aspect ratio (Lr / Wr) of the minimum rectangle RE is less than the threshold Th1 (S330: NO), the confirmation determination unit 234 determines that the selected crack candidate region CC is a crack region that truly represents a crack. It is determined that something that is not a CR and is not a crack is erroneously detected (erroneous detection) (S350). As described above, since the crack generated in the object contains a line segment component, the aspect ratio of the crack region CR that truly represents the crack becomes a value to some extent. Therefore, in the present embodiment, when the aspect ratio of the crack candidate region CC (the minimum rectangular RE including the crack candidate region CC) is relatively large, it is determined that the crack candidate region CC is a true crack region CR, and the crack candidate region CC is determined. When the aspect ratio of (the smallest rectangular RE including) is relatively small, it is determined that the crack candidate region CC is not a true crack region CR.

The candidate confirmation processing unit 230 determines whether or not the selection of all the crack candidate regions CC is completed (S360), and if it is determined that there is an unselected crack candidate region CC (S360: NO), S310. Return to and repeat the above process. On the other hand, when the candidate confirmation processing unit 230 determines that the unselected crack candidate region CC does not exist (S360: YES), the candidate confirmation processing unit 230 ends the crack candidate confirmation processing.

By the crack candidate confirmation process described above, it is confirmed whether or not each crack candidate region CC detected from the target image IO in the crack candidate detection process is a crack region CR that truly represents a crack. As described above, the crack region CR confirmed by the crack candidate confirmation process is shown on the target image IO displayed in the image display area R3 of the processing screen S1 (see FIG. 4).

A-5. Effect of this embodiment:
As described above, the crack detection device 100 of the present embodiment includes a storage unit 110, a line filter processing unit 222, and a candidate determination unit 224. The storage unit 110 stores a line filter group LFG composed of a plurality of line filter LFs. Each line filter LF constituting the line filter group LFG is composed of elements of N rows × N columns (where N is an orthogonal number of 3 or more), and is a straight line passing through a central element CE located in the center of the row direction and the column direction. This is a filter in which the element value of each element overlapping the above is a positive value (for example, 1), and the element value of an element separated from the straight line in a direction orthogonal to the straight line by a predetermined distance or more is 0 or less. The slopes (filter angles) of the straight lines in which the elements whose element values in each line filter LF are positive values (for example, 1) are arranged are different from each other. Further, the line filter processing unit 222 sequentially selects the pixels constituting the target image IO representing the object as the attention pixel SP, and for each line filter LF constituting the line filter group LFG, the line filter LF is set to the attention pixel SP. By arranging the line filter LF on the target image IO so that the central element CE overlaps, and executing a predetermined calculation using the element values of the elements of the line filter LF that overlap each other and the pixel values of the pixels of the target image IO. A line filter that generates at least one processed target image by calculating the original response value and replacing the pixel value of the pixel of interest SP with the response value calculated based on the original response value calculated for each line filter LF. Execute the process. Further, the candidate determination unit 224 sets the pixel whose pixel value is equal to or higher than a predetermined threshold value in the target image IO after the line filter processing to the crack candidate pixel Pcc, which is a pixel representing a crack and has a high probability of being cracked.

As described above, in the crack detection device 100 of the present embodiment, the line filter processing unit 222 executes the above-mentioned line filter processing by using the line filter group LFG. According to such line filtering processing, a region in which pixels having a relatively large pixel value in the target image IO are lined up in a specific direction for a certain length or more (that is, a region having a line segment component having a certain length or more). ), The pixel value of the target image IO after the line filter processing becomes a relatively large value. Further, in the crack detection device 100 of the present embodiment, the candidate determination unit 224 determines a pixel whose pixel value is equal to or greater than a predetermined threshold in the target image IO after the line filter processing, that is, a line segment having a length equal to or longer than a certain level. The pixels constituting the region having the component are set to the crack candidate pixel Pcc. In general, a crack generated in an object contains a component of a line segment having a certain length or more (that is, a component that is elongated and extends linearly). Therefore, according to the crack detection device 100 of the present embodiment, erroneous detection that is not a crack but has a pixel value close to the cracked portion (for example, dirt on the surface of an object) is erroneously detected as a crack, or It is possible to suppress the occurrence of detection omissions in which the cracks that actually exist cannot be detected, and it is possible to improve the accuracy of crack detection using image processing.

Further, in the crack detection device 100 of the present embodiment, the above-mentioned predetermined calculation for calculating the original response value Ro is a power value that should have the pixel value of the target image IO as the base and the element value of the line filter LF as the exponent. This is an operation in which the product of the powers to be calculated for all the elements constituting the line filter LF is used as the original response value Ro. In this way, in a region in which pixels having a relatively large pixel value in the target image IO are lined up in a specific direction for a certain length or more (that is, a region having a line segment component having a certain length or more), a line is formed. The pixel value of the target image IO after the filtering process becomes a large value that can be more clearly distinguished from the pixel values of other regions. Therefore, according to the crack detection device 100 of the present embodiment, it is possible to effectively suppress the occurrence of erroneous detection and detection omission, and it is possible to effectively improve the accuracy of crack detection using image processing. can.

Further, in the crack detection device 100 of the present embodiment, the line filter processing unit 222 generates a processed target image for each line filter LF (for each filter angle). Further, as the response value calculated for generating the target image after processing for the one line filter LF, the original response value Ro calculated by using the one line filter LF is multiplied by the first weighting coefficient W1. The original response value Ro calculated using at least one other line filter LF whose slope of a line (filter angle) is close to that of the one line filter LF is smaller than the first weighting coefficient W1. The weighted response value Rw, which is the sum of each value multiplied by the weighting coefficient W2, is used. Therefore, according to the crack detection device 100 of the present embodiment, the crack can be detected accurately even when the direction of the crack in the target image IO is slightly deviated from the filter angle, and the crack detection using image processing can be performed. The accuracy can be effectively improved.

Further, in the crack detection device 100 of the present embodiment, in the line filter LF, the element value of each element overlapping the straight line passing through the central element CE is the first positive value (for example, 1), and the straight line is changed to the straight line. The element value of the element separated by the first distance or more in the orthogonal direction is 0 or less, and the element value of the element separated from the straight line by less than the first distance in the direction orthogonal to the straight line is the first one. A filter having a second positive value (eg 0.2) that is smaller than the positive value. Therefore, according to the crack detection device 100 of the present embodiment, cracks having a certain width can be detected with high accuracy, and the accuracy of crack detection using image processing can be effectively improved.

Further, in the crack detection device 100 of the present embodiment, the line filter processing unit 222 selects the line filter group LFG in descending order of the value of the number of unit elements N. Therefore, according to the crack detection device 100 of the present embodiment, it is possible to more reliably suppress the occurrence of false detection in which noise pixels are erroneously detected as crack pixels, and the accuracy of crack detection using image processing is effective. Can be improved.

Further, the crack detection device 100 of the present embodiment further includes a display control unit 210 for displaying an image showing a crack candidate pixel (crack candidate pixel Pcc) on the display unit 120. Therefore, according to the crack detection device 100 of the present embodiment, it is possible for the user to grasp the occurrence status of cracks in the object.

Further, in the crack detection device 100 of the present embodiment, the object of the crack detection process is a tiled wall of the building 310, that is, a wall-shaped structure. A wall-shaped structure is an object in which the occurrence of cracks is likely to be a problem, and it is necessary to detect cracks in a relatively wide range. According to the crack detection device 100 of the present embodiment, it is possible to efficiently and accurately detect cracks in such an object.

B. Second embodiment:
FIG. 13 is a flowchart showing the flow of the crack candidate detection process in the second embodiment, FIG. 14 is an explanatory diagram showing an example of the line filter group LFG in the second embodiment, and FIG. 15 is an explanatory diagram showing an example of the line filter group LFG in the second embodiment. It is a flowchart which shows the line filter processing using the line filter group LFG in an embodiment, and FIG. 16 is an explanatory diagram which shows the outline of the line filter processing using the line filter group LFG in the 2nd Embodiment. In the following, among the crack candidate detection processes of the second embodiment, the same processing contents as the crack candidate detection process of the first embodiment described above will be appropriately described by adding the same reference numerals.

As shown in FIG. 14, the line filter group LFG (N) having the number of unit elements N used in the second embodiment is composed of a plurality of line filters LF (N) composed of elements of N rows × N columns. Has been done. Each line filter LF (N) is a filter in which the element value of each element overlapping the straight line passing through the central element CE is a positive value and the element value of the other elements is 0 or less. More specifically, in each line filter LF (N), the element value of each element overlapping the straight line passing through the central element CE is 1, and the element value of the other elements is 0. Note that in FIG. 14, the element value 0 is not shown (the same applies to FIG. 16 and the like).

Further, each of the plurality of line filters LF (N) constituting a certain line filter group LFG (N) has different slopes (filter angles) of the straight lines in which the elements having the element value of 1 are arranged. For example, as shown in FIG. 14, one line filter LF (5) -0 constituting the line filter group LFG (5) having a unit element number N of 5 has an element having an element value of 1 in the horizontal direction ( The other line filter LF (5) -90 is arranged in the horizontal direction (that is, the filter angle is 0 degrees), and the element having the element value of 1 is in the vertical direction (vertical direction in the figure). ) (Ie, the filter angle is 90 degrees). The line filter group LFG (5) also includes a line filter LF (5) -45 having a filter angle of 45 degrees, a line filter LF (5) -135 having a filter angle of 135 degrees, and the like. .. The same applies to other line filter groups LFG.

When the crack candidate detection process (FIG. 13) is started, the line filter group LFG is selected (S210) and the line filter process using the selected line filter group LFG is executed as in the first embodiment. (S220).

When the line filter processing (FIG. 15) is started, the pixel of interest SP is selected (S221), the line filter LF is selected (S222), and the selected line filter LF is the target image, as in the first embodiment. It is arranged on the IO (S223). In column A of FIG. 16, the position of the line filter LF (5) arranged on the target image IO is indicated by a thick solid line.

Next, the line filter processing unit 222 calculates the original response value Ro by executing a predetermined filter calculation using the element values of the elements of the line filter LF that overlap each other and the pixel values of the pixels of the target image IO. (S224). However, in the second embodiment, the filter calculation method at this time is different from the above-described first embodiment. Specifically, the line filter processing unit 222 calculates a multiplication value between the pixel value of the pixel of the target image IO and the element value of the line filter LF, and the multiplication value of the above multiplication values for all the elements constituting the line filter LF. The sum is calculated as the original response value Ro. For example, in the first step of S222 in the line filter processing using the line filter group LFG (5) in which the number of unit elements N is 5, the line filter LF (5) -0 shown in column B of FIG. 16 is selected. It shall be. In this case, the original response value Ro is 1 × 30 + 1 × 150 + 1 × 200 + 1 × 150 + 1 × 30 = 560.

After calculating the original response value Ro (S224), the line filter processing unit 222 determines whether or not all the line filter LFs constituting the currently selected line filter group LFG have been selected (S225). If it is determined that there is an unselected line filter LF (S225: NO), the process returns to S222 and the above processing is repeated. For example, in the second step of S222 in the line filter processing using the line filter group LFG (5) in which the number of unit elements N is 5, the line filter LF (5) -90 shown in column B of FIG. 16 is selected. It shall be. In this case, the line filter LF (5) -90 is arranged on the target image IO (S223), and the original response value Ro is calculated (S224). At this time, as shown in column B of FIG. 16, the original response value Ro is 1 × 200 + 1 × 200 + 1 × 200 + 1 × 200 + 1 × 200 = 1000.

The line filter processing unit 222 repeatedly executes processing from the selection of the line filter LF (S222) to the calculation of the original response value Ro (S224), and all the line filter LFs constituting the currently selected line filter group LFG. (That is, for the currently selected attention pixel SP, the calculation of the original response value Ro for all the line filters LF is completed) (S225: YES), the original response value The maximum response value Rm is calculated based on Ro, and the target image after processing is generated by replacing the pixel value of the attention pixel SP with the maximum response value Rm (S226). The maximum response value Rm corresponds to the response value in the claims. For example, as shown in column B of FIG. 16, when the original response value Ro is calculated for each line filter LF (5) constituting the line filter group LFG (5) in which the number of unit elements N is 5, the line filter LF ( 5) The maximum value (1000) is calculated for -90. Therefore, the line filter processing unit 222 replaces the pixel value of the pixel of interest SP with 1000, which is the original response value Ro calculated for the line filter LF (5) -90 (see column C in FIG. 16).

The fact that the original response value Ro calculated for a certain line filter LF is large means that in the image region overlapping the line filter LF in the target image IO, the direction of the straight line in which the elements having the element value 1 in the line filter LF are lined up. (For example, in the line filter LF (5) -90 shown in column B of FIG. 16, the pixels having relatively large pixel values (that is, close to black) are arranged in the vertical direction (vertical direction in the figure)). means. For example, in the example shown in column A of FIG. 16, in the image region overlapping the line filter LF in the target image IO, pixels having relatively large pixel values are arranged in the vertical direction. Therefore, in this example, the original response value Ro is maximized for the line filter LF (5) -90 in which the elements having the element value of 1 are arranged in the vertical direction. As described above, in the line filter processing of the present embodiment, in the target image IO, a pixel having a relatively large pixel value centered on the pixel of interest SP has a predetermined length (a length corresponding to the number of unit elements N) in a specific direction. ) When they are lined up, the pixel value of the pixel of interest SP is replaced with a relatively large value. On the other hand, in the target image IO, when there is no portion where pixels having a relatively large pixel value are lined up in a specific direction for a predetermined length (length corresponding to the number of unit elements N) or more, centering on the pixel of interest SP. , The pixel value of the pixel of interest SP is replaced with a relatively small value.

The line filter processing unit 222 determines whether or not the selection as the pixel of interest SP is completed for all the pixels constituting the target image IO (S227), and if it is determined that there are unselected pixels (S227). : NO), the process returns to S221 and the above processing is repeated. That is, one of the unselected pixels in the target image IO is newly selected as the pixel of interest SP, and the newly selected pixel of interest SP is subjected to line filter processing using the currently selected line filter group LFG. The pixel value of the pixel of interest SP is replaced with the above-mentioned maximum response value Rm. On the other hand, when it is determined that the selection of all the pixels constituting the target image IO as the pixel of interest SP is completed (S227: YES), the line filter processing unit 222 ends the line filter processing.

As described above, in the line filter processing (FIG. 15) of the present embodiment, in the target image IO, pixels having a relatively large pixel value centered on the sequentially selected pixel of interest SP have a predetermined length (unit) in a specific direction. When there is a place where the pixel value of interest pixel SP is lined up more than (the length corresponding to the number of elements N), the pixel value of the pixel of interest SP is replaced with a relatively large value, and when there is no such place, the pixel value of the pixel of interest SP is Replaced with a relatively small value. Therefore, in the line filter processing of the present embodiment, after the line filter processing, in a region (a region having a line segment component) in which pixels having a relatively large pixel value in the target image IO are lined up in a specific direction for a certain length or more. The pixel value of the target image IO of is relatively large. For example, in column C of FIG. 16, the target image IO after the first line filter processing (the target image IO (1) after the first processing) is shown. In the first processed target image IO (1), the pixel value of the region extending in the vertical direction (vertical direction in the figure) (the region corresponding to the crack region CR) located near the center becomes a relatively large value. There is.

As shown in FIG. 13, when the line filter processing (S220) using the selected line filter group LFG is completed, the line filter processing unit 222 selects all the line filter groups LFG stored in the storage unit 110. It is determined whether or not the process is completed (S230), and if it is determined that there is an unselected line filter group LFG (S230: NO), the process returns to S210 and the above processing is repeated. That is, the line filter processing unit 222 selects a new line filter group LFG and executes line filter processing using the selected line filter group LFG.

When selecting the line filter group LFG (S210), the line filter group LFG having the largest number of unit elements N is selected from the unselected line filter group LFGs. Further, in the second and subsequent line filter processing (S220), the line filter processing is executed for the target image IO after the line filter processing executed immediately before. For example, in the second line filter processing, the line filter processing is executed for the target image IO after the first line filter processing (the target image IO (1) after the first processing). As described above, in the present embodiment, the line filter processing unit 222 selects the line filter group LFG in descending order of the value of the number of unit elements N, and sequentially executes the line filter processing using the selected line filter group LFG. do.

After the selection of the line filter group LFG and the line filter processing using the selected line filter group LFG are repeatedly executed, it is determined that the selection of all the line filter group LFG stored in the storage unit 110 is completed. If (S230: YES), the candidate determination unit 224 has a pixel value in the target image IO at that time (that is, the target image IO after each line filter processing using each line filter group LFG). Pixels that are equal to or greater than a predetermined threshold are set as crack candidate pixels Pcc, which are pixels representing cracks and have a high probability of being cracked (S240). Further, the candidate area setting unit 226 sets the image area including the set crack candidate pixel Pcc as the crack candidate area CC having a high probability of being the image area representing the crack (S250). The method of setting the crack candidate pixel Pcc (S240) and setting the crack candidate region CC (S250) is the same as that of the first embodiment described above.

By the crack candidate detection process described above, the crack candidate region CC including the crack candidate pixel Pcc, which is a pixel representing a crack and has a high probability, is detected from the target image IO.

As described above, in the second embodiment, as in the first embodiment, the crack detection device 100 includes a storage unit 110, a line filter processing unit 222, and a candidate determination unit 224. The storage unit 110 stores a line filter group LFG composed of a plurality of line filter LFs. Each line filter LF constituting the line filter group LFG is composed of elements of N rows × N columns (where N is an orthogonal number of 3 or more), and is a straight line passing through a central element CE located in the center of the row direction and the column direction. In a filter in which the element value of each element overlapping the above is a positive value (for example, 1), and the element value of an element separated from the straight line by a predetermined distance or more in the direction orthogonal to the straight line is 0 or less (for example, 0). be. The slopes (filter angles) of the straight lines in which the elements whose element values in each line filter LF are positive values (for example, 1) are arranged are different from each other. Further, the line filter processing unit 222 sequentially selects the pixels constituting the target image IO representing the object as the attention pixel SP, and for each line filter LF constituting the line filter group LFG, the line filter LF is set to the attention pixel SP. By arranging the line filter LF on the target image IO so that the central element CE overlaps, and executing a predetermined calculation using the element values of the elements of the line filter LF that overlap each other and the pixel values of the pixels of the target image IO. A line filter that generates at least one processed target image by calculating the original response value and replacing the pixel value of the pixel of interest SP with the response value calculated based on the original response value calculated for each line filter LF. Execute the process. Further, the candidate determination unit 224 sets the pixel whose pixel value is equal to or higher than a predetermined threshold value in the target image IO after the line filter processing to the crack candidate pixel Pcc, which is a pixel representing a crack and has a high probability of being cracked.

As described above, in the crack detection device 100 of the present embodiment, the line filter processing unit 222 executes the above-mentioned line filter processing by using the line filter group LFG. According to such line filtering processing, a region in which pixels having a relatively large pixel value in the target image IO are lined up in a specific direction for a certain length or more (that is, a region having a line segment component having a certain length or more). ), The pixel value of the target image IO after the line filter processing becomes a relatively large value. Further, in the crack detection device 100 of the present embodiment, the candidate determination unit 224 determines a pixel whose pixel value is equal to or greater than a predetermined threshold in the target image IO after the line filter processing, that is, a line segment having a length equal to or longer than a certain level. The pixels constituting the region having the component are set to the crack candidate pixel Pcc. In general, a crack generated in an object contains a component of a line segment having a certain length or more (that is, a component that is elongated and extends linearly). Therefore, according to the crack detection device 100 of the present embodiment, erroneous detection that is not a crack but has a pixel value close to the cracked portion (for example, dirt on the surface of an object) is erroneously detected as a crack, or It is possible to suppress the occurrence of detection omissions in which the cracks that actually exist cannot be detected, and it is possible to improve the accuracy of crack detection using image processing.

Further, in the second embodiment, the above-mentioned predetermined operation for calculating the original response value Ro calculates the multiplication value of the pixel value of the target image IO and the element value of the line filter LF, and constitutes the line filter LF. This is an operation in which the sum of the multiplication values for all the elements is the original response value Ro. In this way, in a region in which pixels having a relatively large pixel value in the target image IO are lined up in a specific direction for a certain length or more (that is, a region having a line segment component having a certain length or more), a line is formed. The pixel value of the target image IO after the filtering process becomes a large value that can be more clearly distinguished from the pixel values of other regions. Therefore, according to the crack detection device 100 of the present embodiment, it is possible to effectively suppress the occurrence of erroneous detection and detection omission, and it is possible to effectively improve the accuracy of crack detection using image processing. can.

Further, in the second embodiment, the maximum response value Rm, which is the maximum value among the original response values Ro calculated for each line filter LF, is used as the response value calculated based on the original response value Ro. Therefore, according to the crack detection device 100 of the present embodiment, the crack can be detected with high accuracy, and the accuracy of crack detection using image processing can be effectively improved.

C. Modification example:
The technique disclosed in the present specification is not limited to the above-described embodiment, and can be transformed into various forms without departing from the gist thereof, and for example, the following modifications are also possible.

The configuration of the crack detection device 100 in the above embodiment is merely an example and can be variously modified. For example, the crack detection device 100 may include an imaging unit, and the crack detection processing unit 200 may execute the crack detection processing on an image generated by imaging an object by the imaging unit.

Further, in the above embodiment, the processing screen S1 and the like are displayed on the display unit 120 of the crack detection device 100, but the processing screen S1 and the like are not the display unit 120 provided in the crack detection device 100, but an external display. It may be displayed in. In this case, the crack detection device 100 does not need to include the display unit 120.

Further, in the above embodiment, the type of the line filter group LFG stored in the storage unit 110 can be arbitrarily changed.

Further, in the above embodiment, the number of unit elements N of the line filter group LFG stored in the storage unit 110 can be arbitrarily changed as long as it is an odd number of 3 or more. Further, in the above embodiment, the element values of the line filter LFs constituting the line filter group LFG stored in the storage unit 110 (see FIGS. 6 and 14) are the element values of the elements overlapping the straight line passing through the central element CE. Is a positive value, and can be arbitrarily changed as long as the element value of the element separated from the straight line in the direction orthogonal to the straight line by a predetermined distance or more is 0 or less.

Further, the content of the crack detection process (FIG. 3) in the above embodiment is merely an example and can be variously deformed. For example, in the crack detection process in the above embodiment, the crack candidate confirmation process (S150) is executed, but the crack candidate confirmation process does not necessarily have to be executed. When the crack candidate confirmation process is not executed, the crack candidate region CC detected in the crack candidate detection process (S130) is set as it is in the crack region CR.

Further, in the crack candidate detection process (FIG. 5) in the above embodiment, the plurality of line filter groups LFG are selected in descending order of the number of unit elements N, but the selection order of the line filter group LFG can be arbitrarily changed. Is. Further, in the first embodiment, the weighted response value Rw is adopted as the response value calculated based on the original response value Ro, but the maximum response value Rm may be adopted as in the second embodiment. On the contrary, in the second embodiment, the maximum response value Rm is adopted as the response value calculated based on the original response value Ro, but the weighted response value Rw may be adopted as in the first embodiment. ..

Further, in the above embodiment, the object of the crack detection treatment is a tiled wall of the building 310, but other wall-like structures (for example, a concrete wall or a civil engineering structure of a building). It may be an object other than a wall-like structure.

100: Crack detection device 110: Storage unit 120: Display unit 130: Input unit 140: Interface unit 170: Control unit 190: Bus 200: Crack detection processing unit 210: Display control unit 220: Candidate detection processing unit 222: Line filter processing Part 224: Candidate judgment unit 226: Candidate area setting unit 230: Candidate confirmation processing unit 232: Rectangular calculation unit 234: Confirmation judgment unit 310: Building 320: Imaging device CC: Crack candidate area CE: Central element CP: Crack detection program CR: Crack area IO: Target image Id: Image data LF: Line filter LFG: Line filter group Pb: Background pixel Pc: Crack pixel Pcc: Crack candidate pixel Pn: Noise pixel R1: Image selection area R11: Number display field R2: Method selection area R3: Image display area RE: Minimum rectangle Rm: Maximum response value Ro: Original response value Rw: Weighted response value S1: Processing screen SP: Pixel of interest

Claims (10)

A crack detection device that detects cracks in an object.
It is composed of elements of N rows × N columns (however, N is an orthogonal number of 3 or more), and the element value of each element overlapping the straight line passing through the central element located in the center of the row direction and the column direction is a positive value. A line filter group composed of a plurality of line filters in which the element values of elements separated from the straight line in a direction orthogonal to the straight line by a predetermined distance or more are 0 or less, and the slope of the straight line in each line filter. Are different from each other, with a storage unit that stores the line filter group,
Pixels constituting the target image representing the object are sequentially selected as the pixels of interest, and for each of the line filters constituting the line filter group, the target image is such that the central element of the line filter overlaps the pixels of interest. The original response value is calculated by arranging the line filter on the top and executing a predetermined calculation using the element values of the elements of the line filter overlapping with each other and the pixel values of the pixels of the target image, and the pixel of interest. With a line filter processing unit that executes line filter processing that generates at least one post-processing target image by replacing the pixel value of the above with a response value calculated based on the original response value calculated for each of the line filters. ,
In the processed target image, a candidate determination unit that sets pixels having a pixel value equal to or higher than a predetermined threshold value as crack candidate pixels having a high probability of being cracks.
Equipped with a,
In the predetermined calculation, a power value to be calculated with the pixel value as the base and the element value as a power index is calculated, and the product of the power values for all the elements constituting the line filter is used as the original response value. A crack detection device. The crack detection device according to claim 1.
The line filter processing unit generates a target image after the processing for each line filter.
The response value calculated for generating the processed target image for the one line filter is a value obtained by multiplying the original response value calculated by using the one line filter by a first weighting coefficient. And each of the original response values calculated by using at least one other line filter whose slope of the straight line is close to the one line filter is multiplied by a weighting coefficient smaller than the first weighting factor. A crack detector that is the sum of the values. The crack detection device according to claim 1.
The crack detection device, wherein the response value is the maximum value among the original response values calculated for each of the line filters. The crack detection device according to any one of claims 1 to 3.
In the line filter, the element value of each element overlapping the straight line is the first positive value, and the element value of the element separated from the straight line by the first distance or more in the direction orthogonal to the straight line is 0 or less. A crack detection device, which is a filter in which the element value of an element separated by less than the first distance in a direction orthogonal to the straight line from the straight line is a second positive value smaller than the first positive value. .. The crack detection device according to any one of claims 1 to 4.
The storage unit stores a plurality of the line filter groups having different N values from each other.
The line filter processing unit is executed immediately before selecting one of the plurality of line filter groups and targeting the target image or when the line filter processing is executed at least once. A crack detection device that repeatedly executes the line filter processing using the selected line filter group for the processed target image after the line filter processing. The crack detection device according to claim 5.
The line filter processing unit is a crack detection device that selects the line filter group in descending order of the value of N. The crack detection device according to any one of claims 1 to 6, and further.
A crack detection device including a display control unit that displays an image showing the crack candidate pixels on a display. The crack detection device according to any one of claims 1 to 7.
The object is a crack detection device, which is a wall-shaped structure. It is a crack detection method that detects cracks in an object.
It is composed of elements of N rows × N columns (however, N is an orthogonal number of 3 or more), and the element value of each element overlapping the straight line passing through the central element located in the center of the row direction and the column direction is a positive value. A line filter group composed of a plurality of line filters in which the element values of elements separated from the straight line in a direction orthogonal to the straight line by a predetermined distance or more are 0 or less, and the slope of the straight line in each line filter. Are different from each other, the process of acquiring the line filter group and
Pixels constituting the target image representing the object are sequentially selected as the pixels of interest, and for each of the line filters constituting the line filter group, the target image is such that the central element of the line filter overlaps the pixels of interest. The original response value is calculated by arranging the line filter on the top and executing a predetermined calculation using the element values of the elements of the line filter overlapping with each other and the pixel values of the pixels of the target image, and the pixel of interest. By replacing the pixel value of the above with a response value calculated based on the original response value calculated for each of the line filters, a step of executing a line filter process of generating at least one post-process target image, and a step of executing the line filter process.
In the processed target image, a step of setting a pixel whose pixel value is equal to or higher than a predetermined threshold value as a crack candidate pixel having a high probability of being a pixel representing a crack,
Equipped with a,
In the predetermined calculation, a power value to be calculated with the pixel value as the base and the element value as a power index is calculated, and the product of the power values for all the elements constituting the line filter is used as the original response value. in it, crack detection method. A computer program for detecting cracks in an object
It is composed of elements of N rows × N columns (however, N is an orthogonal number of 3 or more), and the element value of each element overlapping the straight line passing through the central element located in the center of the row direction and the column direction is a positive value. A line filter group composed of a plurality of line filters in which the element values of elements separated from the straight line in a direction orthogonal to the straight line by a predetermined distance or more are 0 or less, and the slope of the straight line in each line filter. Is different from each other, with the function to acquire the line filter group,
Pixels constituting the target image representing the object are sequentially selected as the pixels of interest, and for each of the line filters constituting the line filter group, the target image is such that the central element of the line filter overlaps the pixels of interest. The original response value is calculated by arranging the line filter on the top and executing a predetermined calculation using the element values of the elements of the line filter overlapping with each other and the pixel values of the pixels of the target image, and the pixel of interest. By replacing the pixel value of the above with a response value calculated based on the original response value calculated for each of the line filters, a function of executing a line filter process for generating at least one post-processed target image, and a function of executing the line filter process.
A function of setting a pixel whose pixel value is equal to or higher than a predetermined threshold value in the processed target image as a crack candidate pixel having a high probability of being a pixel representing a crack.
To the computer ,
In the predetermined calculation, a power value to be used as a power index is calculated with the pixel value as the base, and the product of the power values for all the elements constituting the line filter is used as the original response value. Is a computer program.

Images