CN112785587B

CN112785587B - Foreign matter detection method, system, equipment and medium in stacking production process

Info

Publication number: CN112785587B
Application number: CN202110157686.5A
Authority: CN
Inventors: 王永波; 陈伟伟; 许伟; 吕伟
Original assignee: Shanghai Electric Group Corp
Current assignee: Shanghai Electric Group Corp
Priority date: 2021-02-04
Filing date: 2021-02-04
Publication date: 2024-05-31
Anticipated expiration: 2041-02-04
Also published as: CN112785587A

Abstract

The invention discloses a foreign matter detection method, a system, equipment and a medium in a stacking production process, wherein the stacking production process is used for stacking targets with fixed patterns or sizes layer by layer according to preset rules; the foreign matter detection method includes: acquiring a detection image of the layer stacking target when stacking is completed for each layer starting from the second layer; and comparing the characteristics of the detection image with those of a reference image to judge whether the stacking object of the layer has foreign matters, wherein the reference image is an image corresponding to the stacking object of the upper layer. According to the invention, the image corresponding to the stacking object of the previous layer is directly used as the reference image, and the characteristic comparison is carried out with the detection image of the stacking object of the previous layer to judge whether the stacking object of the previous layer has foreign matters or not, and the characteristic comparison is carried out based on the historical data of the previous layer.

Description

Foreign matter detection method, system, equipment and medium in stacking production process

Technical Field

The present invention relates to the field of image processing, and in particular, to a method, a system, an apparatus, and a medium for detecting foreign matters in a stacking production process.

Background

In industrial processes, stacking processes are often involved. The stacking production process refers to stacking or splicing the prepared raw materials or semi-finished products with fixed patterns or sizes layer by layer according to design rules so as to realize the application of a design scheme or a production process. In the case of stacking processes, it is often required that no foreign objects are involved in the process, i.e. that materials or hardware not belonging to the stacking process are not allowed to participate in the stacking, in particular that foreign objects are rejected from being present at the critical contact surfaces.

The current foreign matter detection method for the stacking process mainly includes a sensor device method and an intelligent image processing method. The sensor equipment method is to arrange various sensors such as laser sensing, proximity sensing and the like around a production line of the stacking production process and cooperate with a logic controller to realize foreign matter detection of the stacking process. The intelligent image processing method is used for photographing and detecting each key contact surface in the stacking process through a combination of a picture taking camera and a machine learning algorithm.

The sensor equipment method has low detection time and can realize real-time detection, but the method has relatively limited pertinence, is difficult to apply to foreign matter detection in complex processes, complex environments and large-size stacking production processes, and has high one-time input cost. The intelligent image processing method needs to configure a picture-taking camera with corresponding size and precision according to detection requirements and develop a foreign matter detection algorithm meeting the target so as to adapt to various complex foreign matter detection requirements, and although hardware investment is relatively small, advanced algorithms such as machine learning and the like are needed to be combined, so that software development cost is high, development period is long and efficiency is low.

Disclosure of Invention

The invention aims to overcome the defects of higher software development cost, longer development period and lower efficiency caused by the combination of advanced algorithms such as machine learning and the like in an intelligent image processing method in the prior art, and provides a foreign matter detection method, a foreign matter detection system, foreign matter detection equipment and a foreign matter detection medium in a stacking production process.

The invention solves the technical problems by the following technical scheme:

A foreign matter detection method in a stacking production process, wherein the stacking production process is used for stacking targets with fixed patterns or sizes layer by layer according to preset rules;

the foreign matter detection method includes:

acquiring a detection image of the layer stacking target when stacking is completed for each layer starting from the second layer;

And comparing the characteristics of the detection image with those of a reference image to judge whether the stacking object of the layer has foreign matters, wherein the reference image is an image corresponding to the stacking object of the previous layer.

Preferably, the step of comparing the detected image with a reference image to determine whether the layer stacking object has a foreign object specifically includes:

acquiring bounding box data of the detection image by using a bounding box algorithm;

Judging whether the bounding box data meets a first preset condition or not;

If yes, comparing the characteristics of the detection image with the characteristics of the reference image to judge whether the layer stacking target has foreign matters or not;

if not, carrying out position registration on the detection image and the reference image;

and comparing the registered detection image with the reference image in characteristics to judge whether the layer stacking target has foreign matters or not.

Preferably, the step of determining whether the bounding box data meets a first preset condition specifically includes:

Pre-storing a relation table of bounding boxes and first transfer matrixes, wherein the relation table comprises stacking layer parameters corresponding to each layer, bounding box data corresponding to a detection image and the first transfer matrixes corresponding to the detection image and the reference image;

Judging whether the difference value between the bounding box data and the corresponding bounding box standard data in the relation table is smaller than a threshold value or not;

If yes, performing spatial transposition on the detection image of the layer by using the first transposition matrix in the relation table to obtain a first calibration detection image;

the step of comparing the detected image with the reference image to determine whether the layer stacking object has a foreign object specifically includes:

and comparing the first calibration detection image with the reference image in characteristics to judge whether the layer stacking object has foreign matters or not.

Performing position registration on the detection image and the reference image;

Preferably, the step of performing the position registration on the detection image and the reference image specifically includes:

acquiring first image features of the detection image and second image features of the reference image;

Performing similarity calculation on the first image feature and the second image feature to find matched feature point pairs;

Obtaining image space coordinate transformation parameters according to the characteristic point pairs;

Generating a second transpose matrix according to the image space coordinate transformation parameters;

and completing the position registration of the detection image and the reference image according to the second transpose matrix.

Preferably, the step of comparing the registered detection image with the reference image to determine whether the layer stacking object has a foreign object specifically includes:

If the second transposed matrix does not meet a second preset condition, spatial transposition is carried out on the registered detection images by using the second transposed matrix so as to obtain second calibration detection images;

and comparing the second calibration detection image with the reference image in characteristics to judge whether the layer stacking object has foreign matters or not.

A foreign matter detection system in a stacking production process, wherein the stacking production process is used for stacking targets with fixed patterns or sizes layer by layer according to preset rules;

The foreign matter detection system includes:

a detection image acquisition module for acquiring, for each layer starting from the second layer, a detection image when stacking of the layer stacking target is completed;

And the characteristic comparison module is used for carrying out characteristic comparison on the detection image and a reference image so as to judge whether the stacking object of the layer has foreign matters or not, wherein the reference image is an image corresponding to the stacking object of the upper layer.

Preferably, the feature comparison module specifically includes: the device comprises a data acquisition unit, a judging unit, a first characteristic comparison unit, a second characteristic comparison unit and a first registration unit;

the data acquisition unit is used for acquiring bounding box data of the detection image by using a bounding box algorithm;

The judging unit is used for judging whether the bounding box data meets a first preset condition or not;

if yes, calling the first characteristic comparison unit; if not, calling the first registration unit;

the first feature comparison unit is used for performing feature comparison on the detection image and the reference image to judge whether the layer stacking object has foreign matters or not;

the first registration unit is used for carrying out position registration on the detection image and the reference image;

And the second feature comparison unit is used for comparing the registered detection image with the reference image in a feature mode so as to judge whether the layer stacking target has foreign matters or not.

Preferably, the judging unit is specifically further configured to:

the first feature comparison unit is specifically further configured to:

Preferably, the feature comparison module specifically further includes: a second registration unit and a third feature comparison unit;

The second registration unit is used for carrying out position registration on the detection image and the reference image;

and the third feature comparison unit is used for comparing the registered detection image with the reference image in a feature mode so as to judge whether the layer stacking target has foreign matters or not.

Preferably, the first registration unit is specifically configured to:

Completing the position registration of the detection image and the reference image according to the second transpose matrix;

and/or the number of the groups of groups,

The second feature comparison unit is specifically configured to:

Preferably, the second registration unit is specifically configured to:

and/or the number of the groups of groups,

The third feature comparison unit is specifically configured to:

An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method of foreign object detection in a stack production process of any one of the above when executing the program.

A computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the foreign object detection method in a stack production process of any one of the above.

The invention has the positive progress effects that: according to each layer from the second layer, the image corresponding to the stacking object of the previous layer is directly used as a reference image, and the feature comparison is carried out on the detected image and the reference image when the stacking object of the layer is stacked, so that whether foreign matters exist in the stacking object of the layer or not is judged, and the feature comparison is carried out on the basis of the historical data of the previous layer. Based on the image processing method, the method has better scene suitability.

Drawings

Fig. 1 is a flow chart of a foreign matter detection method in a stacking process according to embodiment 1 of the present invention.

Fig. 2 is a flow chart of a foreign matter detection method in the stacking production process of embodiment 2 of the present invention.

Fig. 3 is a partial flow chart of a foreign matter detection method in a stacking production process according to another embodiment of example 2 of the present invention.

Fig. 4 is a schematic diagram of a foreign matter detection system in a stacking process according to embodiment 3 of the present invention.

Fig. 5 is a schematic diagram of a foreign matter detection system in a stacking process according to embodiment 4 of the present invention.

Fig. 6 is a schematic structural diagram of an electronic device according to embodiment 5 of the present invention.

Detailed Description

The invention is further illustrated by means of the following examples, which are not intended to limit the scope of the invention.

Example 1

The embodiment provides a foreign matter detection method in a stacking production process, wherein the stacking production process is used for stacking targets with fixed patterns or sizes layer by layer according to preset rules; theoretically, the layout after the stacking target of each layer is stacked in the stacking production process is the same.

As shown in fig. 1, the foreign matter detection method includes the steps of:

101. For each layer starting from the second layer, a detection image at the completion of stacking of the layer stack target is acquired. In the stacking production process, the stacking target of the first layer does not detect the foreign matter, and the detection result of the first layer foreign matter detection is default to be free of foreign matter. If a foreign matter actually appears on the surface of the stacking object of the first layer, the foreign matter is detected in the foreign matter detection of the next layer.

102. And comparing the characteristics of the detection image with those of a reference image to judge whether the stacking object of the layer has foreign matters, wherein the reference image is an image corresponding to the stacking object of the upper layer.

According to the foreign matter detection method in the stacking production process, the stacking targets of the first layer do not detect foreign matters, the detection result of the first layer foreign matter detection is defaulted to be free of foreign matters, for each layer starting from the second layer, the detected image of the stacking targets of the layer is compared with the reference image by directly using the image corresponding to the stacking targets of the upper layer as the reference image, so that whether the stacking targets of the layer have foreign matters or not is judged, and the characteristic comparison is carried out based on the historical data of the upper layer; based on the image processing method, the method has better scene suitability.

Example 2

The present embodiment provides a method for detecting a foreign object in a stacking process, which is a further improvement of embodiment 1, specifically, as shown in fig. 2, step 102 specifically includes:

1021. Bounding box data of the detected image is acquired using a bounding box algorithm. It should be noted that, the bounding Box (Bound Box) is an algorithm for solving an optimal bounding space of a discrete point set, and the basic idea is to replace a complex geometric object with a slightly larger volume and simple characteristics, and the bounding Box can quickly acquire the frame and vertex information of the detected image, for example, the stacking target is a rectangular plate, and the bounding Box data includes the vertex information of the rectangular plate.

1022. Judging whether the bounding box data meets a first preset condition, if so, executing step 1023, and if not, executing step 1024;

1023. The detected image is compared with the reference image in characteristics to judge whether the layer stack object has foreign matter.

In an optional implementation manner, the first preset condition includes a first threshold value, by comparing the bounding box data with the first threshold value, when the bounding box data is smaller than the first threshold value, it is indicated that the difference between the positions of the detection image and the reference image is smaller, and the detection image and the reference image can be directly compared in characteristic to determine whether the layer stacking object has foreign objects, so that detection time is saved.

In an alternative embodiment, the first preset condition includes pre-stored bounding box data of the upper platform of the stacking platform and the workpiece of each layer of standard, whether the camera lens has a strict rigid relationship with the stacking line or not can be judged by comparing the bounding box data of the upper platform of the stacking platform and the workpiece obtained after the same stacking step in the same stacking scheme with the pre-stored corresponding bounding box data, if the camera lens has the strict or basically strict rigid relationship with the stacking line, the position registration information in the stacking production process can be reused in the subsequent batch production link, and it is required to be noted that the bounding box data is far less than the position registration, and the position registration step can be omitted when the bounding box data meets the preset condition through the bounding box data judgment, so that the long-time consumption of the position registration step is omitted, and the functions of matching various automatic stacking lines and realizing real-time foreign matter detection are realized.

In an alternative embodiment, step 102 specifically further includes:

1024. the detection image is in position registration with the reference image.

1025. And comparing the registered detection image with a reference image in characteristics to judge whether the layer stacking target has foreign matters or not.

It should be noted that, if the bounding box data does not meet the first preset condition, it is indicated that the detecting camera lens and the stacking line no longer have a strict rigid relationship, or the number of stacked layers is greater and the thickness is greater, so that the camera lens and the stacking line no longer have a strict rigid relationship. At this time, the detection image and the reference image are required to be subjected to position registration, and the registered detection image and the reference image are subjected to feature comparison to judge whether the layer of stacked targets have foreign matters, and the subsequent feature extraction is more accurate through the position registration, so that an accurate data basis is provided for the subsequent foreign matter detection through the feature comparison, and the accuracy of the foreign matter detection is improved.

In an alternative embodiment, step 1022 specifically includes:

If yes, the first transfer matrix in the relation table is used for carrying out space transposition on the detection image of the layer to obtain a first calibration detection image.

Table 1 below shows a schematic diagram of a BT (bounding box-first transition matrix) relationship table of the present embodiment.

TABLE 1

Index numbering	Stacking layer parameters	Bounding box standard data	First transfer matrix
				0x01	1	Mat1	T1
0x02	2	Mat2	T2
				0x03	3	Mat3	T3

In table 1, the stack layer parameters 1, 2, or 3 corresponding to each layer, bounding box standard data Mat1, mat2, mat3 corresponding to the detection image, and first transfer matrices T1, T2, and T3 corresponding to the detection image and the reference image are included.

Step 1023 specifically includes:

the first calibration detection image is compared with the reference image in characteristics to judge whether the layer stacking object has foreign matters or not.

In the scheme, when the bounding box data deviation is smaller or is in an error range, if the position angle of the detection image is inclined, the first calibration detection image is obtained by performing spatial transposition on the detection image through a first transposition matrix in the relation table, and the first calibration detection image is subjected to characteristic comparison with the reference image to judge whether the layer of stacked targets have foreign matters or not, so that the accuracy of foreign matter detection is improved.

In an alternative embodiment, the foreign matter detection method in the stacking production process further includes the steps of:

103. and if the layer stacking target has foreign matters, sending alarm information. Through sending alarm information when detecting that stacking target exists the foreign matter, can conveniently fix a position that the foreign matter appears, in time carry out abnormal handling, guarantee the safe operation of stacking production line.

In an alternative embodiment, as shown in fig. 3, step 102 specifically includes:

1026. performing position registration on the detection image and the reference image;

1027. and comparing the registered detection image with a reference image in characteristics to judge whether the layer stacking target has foreign matters or not.

In the scheme, the detection image of each layer from the second layer is subjected to position registration with the reference image, so that the bounding box judging step is omitted, the registered detection image is subjected to feature comparison with the reference image to judge whether the stacked targets of the layers have foreign matters, and the subsequent feature extraction is more accurate through the position registration, so that an accurate data basis is provided for the subsequent foreign matter detection through the feature comparison, and the accuracy of the foreign matter detection is improved. Although each layer carries out position registration, compared with the existing machine learning algorithm, the software development cost is still low, and the application range is wider. The transposed matrix of the position registration results are smaller or the space transposition step of foreign matter detection can be omitted when the acceptable error range of the stacking production line is met, so that the detection time is further reduced.

In an alternative embodiment, step 1024 or step 1026 specifically includes:

acquiring first image features of a detection image and second image features of a reference image;

Obtaining image space coordinate transformation parameters according to the feature point pairs;

And finishing the position registration of the detection image and the reference image according to the second transpose matrix.

According to the image position registration method, on the basis of image feature extraction, the image space coordinate transformation parameters are obtained according to the feature point pairs after the matched feature point pairs are found through similarity measurement, the image space coordinate transformation parameters are expressed as the first transformation matrix in an algorithm, the position registration of the detection image and the reference image can be completed through the first transformation matrix, the subsequent feature extraction is more accurate through the position registration, an accurate data basis is provided for the subsequent foreign object detection through feature comparison, and the accuracy of foreign object detection is improved.

In an alternative embodiment, step 1027 specifically includes:

if the second transposed matrix does not meet the second preset condition, spatial transposition is performed on the registered detection images of the stacking targets by using the second transposed matrix to obtain a second calibration detection image. Specifically, the second preset condition includes an error threshold acceptable to the stacking line, where the error threshold is an angle value. For example, as the stacking operation proceeds, the detected image may be tilted, if the detected image is compared with the reference image, and the tilt angle of the detected image is larger, for example, 5 degrees, and at this time, the tilt angle is larger than the error threshold value acceptable by the stacking line by 1 degree, then the second transposed matrix is used to spatially transpose the registered detected image of the stacking target to obtain a second calibration detected image, where the second calibration detected image is not angularly tilted compared with the reference image.

In the scheme, the second transposed matrix obtained by the position registration is larger in result and does not accord with the situation that the second transposed matrix obtained by the position registration is used for carrying out space transposition on the detection image to obtain a second calibration detection image when the second transposed matrix is not in the acceptable error range of the stacking production line, and the second calibration detection image is compared with the reference image in characteristics to judge whether the stacking target of the layer has foreign matters or not, so that the accuracy of foreign matter detection is improved.

The implementation steps of the foreign matter detection method in the stacking production process of the embodiment include: preparing a partial step and implementing a partial step.

Wherein, the preparation part comprises the following steps:

1) The result of the first layer foreign matter detection is defaulted to be foreign matter-free.

2) Registration reference information is prepared for foreign object detection of the second layer and later layers. Specifically, a BT (bounding box-first transpose matrix) relationship table is created, and after the index numbers for reference are filled, and stack layer parameters are supplemented, the stack layer that needs to perform foreign object detection needs to store bounding box data and transpose matrices in the BT relationship table. In the stacking production line debugging process, image acquisition is performed on each stacking layer, and bounding box data of a stacking platform are calculated. And performing registration calculation on the position of the current stacking layer by using the image corresponding to the stacking target of the upper layer as a reference image, and completing the BT relation table.

The implementation part comprises the following steps:

1) Foreign matter detection of the first stacked layer (0 x 01) is performed. The layer is defaulted to be free of foreign matter.

2) Foreign matter detection is performed on the next stacked layer (0 x 02). And acquiring a stacked layer image acquisition result as a detection image, using a first stacked layer image without foreign matters as a reference image, executing bounding box judgment, selecting a corresponding transpose matrix to transpose the detection image if the bounding box data difference is small, and performing feature comparison and result judgment. If the bounding box data differ significantly, a conventional foreign object detection procedure of position registration and feature comparison is performed.

3) Foreign matter detection is performed on the nth stacked layer (0 xN). And acquiring an N-th stacking layer image acquisition result as a detection image, using the (N-1) -th stacking layer result without foreign matters as a reference image, executing bounding box judgment, if the bounding box data difference is small, selecting a corresponding transpose matrix to transpose the detection image of the stacking result, and carrying out feature comparison and result judgment. If the bounding box data differ significantly, a conventional foreign object detection procedure of position registration and feature comparison is performed. The data in the table is updated or not updated according to the actual use condition of the BT relation table.

The following is a specific application of the foreign matter detection method in the stack production process by way of example: the silicon steel sheet stacking production line of a certain generator factory uses a robot to automatically stack silicon steel sheets, the total stacking area is about 8 square meters, the average thickness of the silicon steel sheets is 2 millimeters, and the single stacking height is about 50 centimeters. The stacking process requires strictly no foreign matter intervention.

In the stacking production line foreign matter detection process, a high-definition camera is used for foreign matter intrusion detection after each stacking operation of the robot. The stacking workbench is fastened on the ground, and the high-definition camera is fastened on the wall surface, namely, the camera has a strict rigid relation with the stacking operation area and cannot change during normal operation.

(1) During production line debugging, capturing and storing a first layer stacking effect without foreign matters, and calculating and storing bounding boxes, registration information and corresponding relations of the bounding boxes and registration information after each stacking step is executed in the complete stacking process, namely a relation table of the bounding boxes and a first transfer matrix.

(2) During production line production, the foreign matter detection flow cooperates with robot stacking operation, after the first layer of the process robot is stacked, the foreign matter detection flow uses pre-stored detection references to detect the foreign matters, and then each layer uses an image corresponding to a stacking target without the foreign matters detected in the previous time as a reference image. When detecting foreign matters, firstly, carrying out bounding box relation judgment, when the change of bounding box data is small, selecting a corresponding transposed matrix by using a pre-stored bounding box-first transposed matrix relation table, and carrying out characteristic comparison after carrying out space transposition on a detected image so as to judge whether the layer of stacked targets have foreign matters or not.

(3) The rigid relationship between the detection camera and the stacking table basically ensures the usability of the relationship table of the bounding box-first transfer matrix, and the use of the bounding box greatly reduces the time of the foreign matter judging process.

The foreign matter detection method in the stacking production process has better adaptability than the traditional sensor method, and meets the foreign matter detection requirements of more scenes; compared with an intelligent image processing method, the method has lower development cost and can realize real-time detection.

Example 3

The embodiment provides a foreign matter detection system in a stacking production process, wherein the stacking production process is used for stacking targets with fixed patterns or sizes layer by layer according to preset rules; theoretically, the layout after the stacking target of each layer is stacked in the stacking production process is the same.

As shown in fig. 4, the foreign matter detection system includes the following modules:

A detection image acquisition module 1 for acquiring, for each layer starting from the second layer, a detection image at the completion of stacking of the layer stack target. In the stacking production process, the stacking target of the first layer does not detect the foreign matter, and the detection result of the first layer foreign matter detection is default to be free of foreign matter. If a foreign matter actually appears on the surface of the stacking object of the first layer, the foreign matter is detected in the foreign matter detection of the next layer.

And the feature comparison module 2 is used for comparing the features of the detected image with the features of a reference image so as to judge whether the stacking object of the layer has foreign matters, wherein the reference image is an image corresponding to the stacking object of the upper layer.

In the embodiment, the foreign matter detection system needle in the stacking production process of the first layer performs no foreign matter detection on the stacking target of the first layer, defaults that the detection result of the first layer foreign matter detection is foreign matter-free, and for each layer starting from the second layer, the detected image of the stacking target of the layer when stacking is completed is compared with the reference image by directly using the image corresponding to the stacking target of the upper layer as the reference image so as to judge whether the stacking target of the layer has foreign matter or not, and the characteristic comparison is performed based on the historical data of the upper layer. Based on the image processing method, the method has better scene suitability.

Example 4

The present embodiment provides a foreign matter detection system in a stacking production process, which is a further improvement of embodiment 3, specifically, as shown in fig. 5, the feature contrast module 2 specifically includes: a data acquisition unit 21, a judgment unit 22, a first feature comparison unit 23, a first registration unit 24, and a second feature comparison unit 25;

A data acquisition unit 21 for acquiring bounding box data of the detection image using a bounding box algorithm. It should be noted that, the bounding Box (Bound Box) is an algorithm for solving an optimal bounding space of a discrete point set, and the basic idea is to replace a complex geometric object with a slightly larger volume and simple characteristics, and the bounding Box can quickly acquire the frame and vertex information of the detected image, for example, the stacking target is a rectangular plate, and the bounding Box data includes the vertex information of the rectangular plate.

A judging unit 22, configured to judge whether the bounding box data meets a first preset condition, and if yes, invoke the first feature comparison unit 23; if not, the first registration unit 24 is invoked;

a first feature comparing unit 23 for feature comparing the detected image with the reference image to determine whether or not the layer stack object has foreign matter.

In an alternative embodiment, the feature comparison module 2 specifically further includes:

If the bounding box data does not meet the first preset condition, invoking the first registration unit 24;

a first registration unit 24 for performing a position registration of the detection image and the reference image.

And a second feature comparing unit 25 for feature comparing the registered detection image with the reference image to determine whether or not the layer stack object has foreign matter.

In an alternative embodiment, the judging unit 22 is specifically further configured to:

TABLE 1

The first feature comparison unit 23 is specifically further configured to:

In an alternative embodiment, the foreign object detection system in the stack production process further comprises the following modules:

And the alarm module 3 is used for sending alarm information if the layer stacking target has foreign matters. Through sending alarm information when detecting that stacking target exists the foreign matter, can conveniently fix a position that the foreign matter appears, in time carry out abnormal handling, guarantee the safe operation of stacking production line.

And the third feature comparison unit is used for performing feature comparison on the registered detection image and the reference image to judge whether the layer stacking target has foreign matters or not.

In an alternative further embodiment, the first registration unit 24 is specifically configured to:

In an alternative further embodiment, the second registration unit is specifically configured to:

In an alternative embodiment, the second feature comparison unit 25 is specifically configured to:

In an alternative embodiment, the third feature comparison unit is specifically configured to:

The implementation process of the foreign matter detection system in the stacking production process of the embodiment includes: the preparation part and the implementation part.

Wherein, the preparation part is as follows:

The implementation part is as follows:

4) Foreign matter detection of the first stacked layer (0 x 01) is performed. The layer is defaulted to be free of foreign matter.

5) Foreign matter detection is performed on the next stacked layer (0 x 02). And acquiring a stacked layer image acquisition result as a detection image, using a first stacked layer image without foreign matters as a reference image, executing bounding box judgment, selecting a corresponding transpose matrix to transpose the detection image if the bounding box data difference is small, and performing feature comparison and result judgment. If the bounding box data differ significantly, a conventional foreign object detection procedure of position registration and feature comparison is performed.

6) Foreign matter detection is performed on the nth stacked layer (0 xN). And acquiring an N-th stacking layer image acquisition result as a detection image, using the (N-1) -th stacking layer result without foreign matters as a reference image, executing bounding box judgment, if the bounding box data difference is small, selecting a corresponding transpose matrix to transpose the detection image of the stacking result, and carrying out feature comparison and result judgment. If the bounding box data differ significantly, a conventional foreign object detection procedure of position registration and feature comparison is performed. The data in the table is updated or not updated according to the actual use condition of the BT relation table.

The following is a specific application of the foreign matter detection system in the stack production process of the present embodiment by way of example: the silicon steel sheet stacking production line of a certain generator factory uses a robot to automatically stack silicon steel sheets, the total stacking area is about 8 square meters, the average thickness of the silicon steel sheets is 2 millimeters, and the single stacking height is about 50 centimeters. The stacking process requires strictly no foreign matter intervention.

The foreign matter detection system in the stacking production process of the embodiment has better adaptability than the traditional sensor system, and meets the foreign matter detection requirements of more scenes; compared with an intelligent image processing system, the system has lower development cost and can realize real-time detection.

Example 5

Fig. 6 is a schematic structural diagram of an electronic device according to embodiment 5 of the present invention. The electronic device includes a memory, a processor, and a computer program stored on the memory and executable on the processor, which when executed implements the foreign matter detection method in the stack production process of embodiment 1 or 2. The electronic device 30 shown in fig. 6 is merely an example, and should not be construed as limiting the functionality and scope of use of embodiments of the present invention.

As shown in fig. 6, the electronic device 30 may be embodied in the form of a general purpose computing device, which may be a server device, for example. Components of electronic device 30 may include, but are not limited to: the at least one processor 31, the at least one memory 32, a bus 33 connecting the different system components, including the memory 32 and the processor 31.

The bus 33 includes a data bus, an address bus, and a control bus.

Memory 32 may include volatile memory such as Random Access Memory (RAM) 321 and/or cache memory 322, and may further include Read Only Memory (ROM) 323.

Memory 32 may also include a program/utility 325 having a set (at least one) of program modules 324, such program modules 324 including, but not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of which may include an implementation of a network environment.

The processor 31 executes various functional applications and data processing, such as a foreign matter detection method in the stack production process of embodiment 1 or 2 of the present invention, by running a computer program stored in the memory 32.

The electronic device 30 may also communicate with one or more external devices 34 (e.g., keyboard, pointing device, etc.). Such communication may be through an input/output (I/O) interface 35. Also, model-generating device 30 may also communicate with one or more networks, such as a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the internet, via network adapter 36. As shown, network adapter 36 communicates with the other modules of model-generating device 30 via bus 33. It should be appreciated that although not shown in the figures, other hardware and/or software modules may be used in connection with the model-generating device 30, including, but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (disk array) systems, tape drives, data backup storage systems, and the like.

It should be noted that although several units/modules or sub-units/modules of an electronic device are mentioned in the above detailed description, such a division is merely exemplary and not mandatory. Indeed, the features and functionality of two or more units/modules described above may be embodied in one unit/module in accordance with embodiments of the present invention. Conversely, the features and functions of one unit/module described above may be further divided into ones that are embodied by a plurality of units/modules.

Example 6

The present embodiment provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the foreign matter detection method steps in the stack production process of embodiment 1 or 2.

More specifically, among others, readable storage media may be employed including, but not limited to: portable disk, hard disk, random access memory, read only memory, erasable programmable read only memory, optical storage device, magnetic storage device, or any suitable combination of the foregoing.

In a possible embodiment, the invention may also be implemented in the form of a program product comprising program code for causing a terminal device to carry out the steps of the foreign object detection method in a stack production process implementing embodiment 1 or 2, when said program product is run on the terminal device.

Wherein the program code for carrying out the invention may be written in any combination of one or more programming languages, which program code may execute entirely on the user device, partly on the user device, as a stand-alone software package, partly on the user device and partly on the remote device or entirely on the remote device.

While specific embodiments of the invention have been described above, it will be appreciated by those skilled in the art that this is by way of example only, and the scope of the invention is defined by the appended claims. Various changes and modifications to these embodiments may be made by those skilled in the art without departing from the principles and spirit of the invention, but such changes and modifications fall within the scope of the invention.

Claims

1. A foreign matter detection method in a stacking production process, wherein the stacking production process is used for stacking targets with fixed patterns or sizes layer by layer according to preset rules;

The foreign matter detection method is characterized by comprising the following steps:

Judging whether the bounding box data meets a first preset condition or not;

if yes, comparing the characteristics of the detection image with those of a reference image to judge whether the stacking object of the layer has foreign matters or not, wherein the reference image is an image corresponding to the stacking object of the upper layer;

Performing feature contrast on the registered detection image and the reference image to judge whether foreign matters exist in the layer stacking target or not;

The step of judging whether the bounding box data meets a first preset condition specifically comprises the following steps:

2. A foreign matter detection method in a stacking production process, wherein the stacking production process is used for stacking targets with fixed patterns or sizes layer by layer according to preset rules;

Judging whether the bounding box data meets a first preset condition or not;

the step of performing position registration on the detection image and the reference image specifically includes:

3. A foreign matter detection method in a stacking production process, wherein the stacking production process is used for stacking targets with fixed patterns or sizes layer by layer according to preset rules;

Performing position registration on the detection image and a reference image, wherein the reference image is an image corresponding to a stacking target of the upper layer;

4. The foreign object detection method in the stack production process according to claim 2 or 3, wherein the step of comparing the registered detection image with the reference image to determine whether the layer stacking object has a foreign object specifically includes:

5. A foreign matter detection system in a stacking production process, wherein the stacking production process is used for stacking targets with fixed patterns or sizes layer by layer according to preset rules;

characterized in that the foreign matter detection system includes:

The feature comparison module is used for comparing the features of the detection image with a reference image to judge whether the stacking object of the layer has foreign matters or not, wherein the reference image is an image corresponding to the stacking object of the upper layer;

the characteristic comparison module specifically comprises: the device comprises a data acquisition unit, a judging unit, a first characteristic comparison unit, a second characteristic comparison unit and a first registration unit;

The second feature comparison unit is used for performing feature comparison on the registered detection image and the reference image to judge whether the layer stacking target has foreign matters or not;

the judging unit is specifically further configured to:

the first feature comparison unit is specifically further configured to:

6. A foreign matter detection system in a stacking production process, wherein the stacking production process is used for stacking targets with fixed patterns or sizes layer by layer according to preset rules;

characterized in that the foreign matter detection system includes:

the first registration unit is specifically configured to:

the second feature comparison unit is specifically configured to:

7. A foreign matter detection system in a stacking production process, wherein the stacking production process is used for stacking targets with fixed patterns or sizes layer by layer according to preset rules;

characterized in that the foreign matter detection system includes:

the characteristic comparison module specifically further comprises: a second registration unit and a third feature comparison unit;

the third feature comparison unit is used for performing feature comparison on the registered detection image and the reference image to judge whether the layer stacking target has foreign matters or not;

the second registration unit is specifically configured to:

The third feature comparison unit is specifically configured to:

if the second transposed matrix does not meet a second preset condition, spatial transposition is performed on the registered detection images by using the second transposed matrix to obtain second calibration detection images.

8. The foreign object detection system in a stack production process of claim 7, wherein:

9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method for detecting foreign objects in a stack production process according to any one of claims 1 to 4 when executing the program.

10. A computer-readable storage medium, on which a computer program is stored, characterized in that the program, when executed by a processor, implements the steps of the foreign matter detection method in a stack production process as claimed in any one of claims 1 to 4.