JP2019086914A - Article inspection device, article inspection method and article inspection program - Google Patents

Abstract

To improve inspection accuracy in an appearance inspection of an article.SOLUTION: An inspection apparatus includes: calculation means for classifying pixels of each color included in a predetermined area in an inspection target article image obtained by photographing an inspection target article and calculating the number of pixels of each color; comparison means for comparing the number of the calculated pixels of each color with reference to the number of pixels of each color included in an area corresponding to the predetermined area in a model image; and determination means for determining whether or not, based on the comparison result, the inspection target article passes an inspection.SELECTED DRAWING: Figure 3

Description

  The present invention relates to an inspection apparatus, an inspection method, and an inspection program.

  Conventionally, in the appearance inspection at the time of shipment of a specific article (for example, a commodity such as a lunch box, bread, vegetables, fruits, flowers, etc.), a judgment of success or failure is made by paying attention to the color of the appearance of the article. There is. For this reason, in an inspection apparatus which performs an appearance inspection of the article, for example, an inspection image obtained by imaging the article is matched with a predetermined model image (an image obtained by imaging an article determined to be acceptable in advance). The decision is being made.

JP, 2005-4612, A Japanese Patent Publication No. 2004-502250

  However, in the case of an article as described above, the color of the appearance is not uniform even if the article is determined to pass the appearance inspection. For example, there is a certain degree of individual difference in appearance color such as uneven baking of a bread or a pattern of a flower, and the same color is not completely reproduced. For this reason, the inspection method for matching with a predetermined model image has a limit in inspection accuracy.

  In one aspect, it is an object of the present invention to improve inspection accuracy in visual inspection of articles.

According to one aspect, the inspection device
Calculation means for classifying the pixels included in a predetermined area according to colors and calculating the number of pixels of each color in the inspection image obtained by photographing the article to be inspected;
A comparison unit that refers to the number of pixels of each color of the pixels included in the area corresponding to the predetermined area in the model image, and compares the calculated number of pixels of each color;
And determination means for determining whether the item to be inspected is acceptable or not based on the comparison result.

  In the appearance inspection of an article, it is possible to improve the inspection accuracy.

It is a figure showing an example of system configuration of an inspection system. It is a figure which shows an example of the hardware constitutions of the inspection apparatus. It is a figure which shows an example of a function structure of the production | generation part of the inspection apparatus. It is a figure which shows the specific example of area | region division processing of a model image. It is a figure which shows the specific example of a compression process. It is a figure which shows the specific example of a RGB value conversion process. It is a figure which shows the specific example of a model data calculation process. It is a figure which shows the specific example of a model data selection process. It is a figure which shows the example of model data. It is a figure which shows an example of a function structure of the inspection part of the inspection apparatus. It is a flowchart which shows the flow of determination processing. It is a figure which shows the specific example of determination processing. It is a figure for demonstrating the effect of the inspection apparatus.

  Hereinafter, each embodiment will be described with reference to the attached drawings. In the present specification and the drawings, components having substantially the same functional configuration will be denoted by the same reference numerals and redundant description will be omitted.

First Embodiment
<System configuration of inspection system>
First, a system configuration of an inspection system including the inspection device according to the first embodiment will be described. FIG. 1 is a diagram showing an example of a system configuration of an inspection system.

  As shown in FIG. 1, the inspection system 100 includes an imaging device 110 and an inspection device 120. The imaging device 110 and the inspection device 120 are connected via a wired or wireless connection.

  The imaging device 110 transmits, to the inspection device 120, image data including a model image obtained by capturing in advance an article determined to be a pass. In addition, the imaging device 110 transmits, to the inspection device 120, image data including an inspection image obtained by imaging the articles 141 to 143 to be inspected. In the first embodiment, the articles 141 to 143 to be inspected are “alarm lunches” and are conveyed by the conveyance device 130 to the imaging range of the imaging device 110.

  The inspection device 120 is an example of an image processing device that processes image data transmitted from the imaging device 110. A generation program and an inspection program are installed in the inspection device 120, and when the program is executed, the inspection device 120 functions as a generation unit 121 and an inspection unit 122.

  The generation unit 121 generates model data based on a model image included in the image data. Further, the generation unit 121 stores the generated model data in the model storage unit 123.

  The inspection unit 122 generates inspection data for determining pass / fail of the article to be inspected based on the inspection image included in the image data. Further, the inspection unit 122 reads out model data corresponding to the generated inspection data from the model storage unit 123, and compares the generated inspection data with the generated inspection data to determine whether the product to be inspected is acceptable or not.

<Hardware Configuration of Inspection Device>
Next, the hardware configuration of the inspection device 120 will be described. FIG. 2 is a diagram showing an example of the hardware configuration of the inspection device. As shown in FIG. 2, the inspection device 120 includes a central processing unit (CPU) 201, a read only memory (ROM) 202, and a random access memory (RAM) 203. The CPU 201, the ROM 202, and the RAM 203 form a so-called computer.

  Further, the inspection device 120 includes an auxiliary storage device 204, an operation device 205, a display device 206, an I / F (Interface) device 207, and a drive device 208. The hardware of the inspection device 120 is mutually connected via the bus 209.

  The CPU 201 is an arithmetic device that executes various programs (for example, a generation program, an inspection program, and the like) installed in the auxiliary storage device 204.

  The ROM 202 is a non-volatile memory. The ROM 202 functions as a main storage device that stores various programs, data, and the like necessary for the CPU 201 to execute various programs installed in the auxiliary storage device 204. Specifically, the ROM 202 stores a boot program such as BIOS (Basic Input / Output System) or EFI (Extensible Firmware Interface).

  The RAM 203 is a volatile memory such as a dynamic random access memory (DRAM) or a static random access memory (SRAM). The RAM 203 functions as a main storage device which provides a work area to be expanded when the various programs installed in the auxiliary storage device 204 are executed by the CPU 201.

  The auxiliary storage device 204 is an auxiliary storage device that stores various programs, data generated by executing the various programs, and the like. For example, the model storage unit 123 is realized in the auxiliary storage device 204.

  The operation device 205 is an input device for the administrator of the inspection device 120 to input various instructions to the inspection device 120. The display device 206 is a display device that displays the internal state of the inspection device 120 and the like. The I / F device 207 is a connection device for connecting to the imaging device 110.

  The drive device 208 is a device for setting the recording medium 210. The recording medium 210 mentioned here includes a medium for optically, electrically or magnetically recording information, such as a CD-ROM, a flexible disk, a magneto-optical disk or the like. The recording medium 210 may also include a semiconductor memory or the like that electrically records information, such as a ROM, a flash memory, or the like.

  The drive device 208 may be used when installing various programs in the auxiliary storage device 204. Specifically, the administrator of the inspection apparatus 120 sets the distributed recording medium 210 in the drive apparatus 208, and the drive apparatus 208 reads out various programs recorded in the recording medium 210, thereby the auxiliary storage apparatus 204. Various programs may be installed on

<Functional Configuration of Generation Unit of Inspection Device>
Next, the functional configuration of the generation unit 121 of the inspection device 120 will be described. FIG. 3 is a diagram illustrating an example of a functional configuration of a generation unit of the inspection device. As illustrated in FIG. 3, the generation unit 121 includes an image data acquisition unit 301, an area division unit 302, a compression unit 303, an RGB value conversion unit 304, a model data calculation unit 305, and a selection unit 306.

  The image data acquisition unit 301 acquires image data from the imaging device 110 and extracts a model image. Further, the image data acquisition unit 301 notifies the region division unit 302 of the extracted model image.

  The area dividing unit 302 divides the model image into a plurality of predetermined areas. Also, the area division unit 302 notifies the compression unit 303 of the area image of each area obtained by the division.

  The compression unit 303 compresses each area image by calculating a pixel value (RGB value) of 1 for each area image by grouping pixel areas of a predetermined number of pixels into one pixel for each area image. . Also, the compression unit 303 notifies the RGB value conversion unit 304 of each compressed region image.

  The RGB value conversion unit 304 performs color conversion on the RGB values of each pixel of each compressed region image. Also, the RGB value conversion unit 304 notifies the model data calculation unit 305 of each of the color-converted area images.

  The model data calculation unit 305 classifies the pixels included in the area image for each color for each area image after color conversion, and calculates the number of pixels of each color to calculate model data candidates for each area. The model data calculation unit 305 also notifies the selection unit 306 of the calculated model data candidates.

  The selecting unit 306 classifies the calculated model data candidates for each area, and extracts a set of model data candidates similar to each other from the classified model data candidates. Further, the selection unit 306 selects one model data candidate from each of the sets of model data candidates similar to each other, and registers the selected model data candidate in the model storage unit 123 as model data.

<Specific Example of Processing of Each Unit of Generation Unit of Inspection Device>
Next, a specific example of the process of each unit (each unit excluding the image data acquisition unit 301) of the generation unit 121 of the inspection device 120 will be described.

(1) Specific Example of Region Division Processing by Region Division Unit 302 First, a specific example of region division processing of a model image by the region division unit 302 of the generation unit 121 will be described. FIG. 4 is a diagram showing a specific example of region segmentation processing of a model image. As shown in FIG. 4A, the model image 400 obtained by photographing in advance an article (a lunch box) determined to be acceptable includes a boundary portion 400BL indicating the outer edge of the lunch box container and the partition inside the lunch box.

  According to the model image 400, the inside of the container of the lunchbox is divided into four storage units (storage units 401, 402, 403, and 404) by the boundary 400BL. Further, according to the model image 400, the accommodation portions 401 to 403 accommodate one kind of contents (contents 411, 412, 413). The contents 411 are, for example, “rice”, the contents 412 are, for example, a “salad”, and the contents 413 are, for example, “pickles”.

  Furthermore, according to the model image 400, the storage unit 404 stores two types of contents (contents 414a and 414b). The contents 414 a are, for example, “hamburgers”, and the contents 414 b are, for example, “fried potatoes”.

  In the first embodiment, the region division unit 302 divides the model image 400 for each type of content. FIG. 4B is a diagram showing division positions when the model image 400 is divided for each type of content.

  As shown in FIG. 4B, the area division unit 302 divides the area corresponding to the storage section 401 in which the content 411 is stored from the model image 400, and sets an area image 421 (area name = "area 1". Extract). In addition, the region division unit 302 divides the region corresponding to the storage unit 402 in which the content 412 is stored from the model image 400, and extracts the region image 422 (referred to as region name = "region 2"). In addition, the area division unit 302 divides an area corresponding to the storage unit 403 in which the content 413 is stored from the model image 400, and extracts an area image 423 (referred to as area name = "area 3").

  In addition, the area division unit 302 divides the left area of the accommodation unit 404 in which the content 414 a is accommodated from the model image, and extracts an area image 424 (referred to as area name = “area 4”). Furthermore, the right side area of the accommodation unit 404 in which the contents 414 b are accommodated is divided from the model image, and an area image 425 (area name = referred to as “area 5”) is extracted.

(2) Specific Example of Compression Processing by Compression Unit 303 Next, a specific example of compression processing by the compression unit 303 of the generation unit 121 will be described. FIG. 5 is a diagram showing a specific example of the compression process.

  As shown in FIG. 5, the compression unit 303 extracts and extracts a pixel area of a predetermined number of pixels (for example, a pixel area consisting of a total of 9 pixels of 3 pixels × 3 pixels) in each of the area images 421 to 425 The average value of the pixel values of each pixel included in the selected pixel area is calculated. Thus, the compression unit 303 compresses the extracted pixel area into one pixel having the calculated pixel value of one.

  The example of FIG. 5 shows a case where “210” is calculated as the average value of the pixel values of the respective pixels included in the pixel area 501 of the area image 421, and is set as the pixel value of one pixel after compression. The compression unit 303 executes such compression processing on the entire range in the region image for each of the region images 421 to 425. As a result, the compression unit 303 can reduce the amount of data of each of the region images 421 to 425.

(3) Specific Example of RGB Value Conversion Processing by RGB Value Conversion Unit 304 Next, a specific example of RGB value conversion processing by the RGB value conversion unit 304 will be described. FIG. 6 is a diagram showing a specific example of the RGB value conversion process.

The RGB value conversion unit 304 performs color conversion on the RGB values of each pixel of each compressed region image. In this embodiment, for example, an R value of 256 gradations of 0 to 255 is divided into 11 gradation ranges, and color conversion of R values belonging to each gradation range is performed to a representative value of each gradation range. . The example of FIG.
Color conversion of an R value belonging to a gradation range of 0 to 24 into a representative value of 0,
Color conversion of an R value belonging to a gradation range of 25 to 49 into a representative value of 25;
Color conversion of an R value belonging to a gradation range of 50 to 74 into a representative value of 50,
Color conversion of R value belonging to gradation range = 75 to 99 to representative value = 75,
Color conversion of an R value belonging to the gradation range of 100 to 124 to a representative value of 100,
Color conversion of an R value belonging to a gradation range of 125 to 149 into a representative value of 125,
And so on. As a result, the R value of 256 gradations is color converted to a representative value of 11.

  Similarly, the RGB value conversion unit 304 divides, for example, the G value of 256 gradations of 0 to 255 into 11 gradation ranges, and represents the G values belonging to each gradation range as a representative of each gradation range. Convert color to a value. As a result, the G value of 256 gradations is color converted to the representative value of 11.

  Similarly, the RGB value conversion unit 304 divides the B value of 256 gradations of 0 to 255, for example, into 11 gradation ranges, and represents the B values belonging to each gradation range as a representative of each gradation range. Convert color to a value. As a result, the B value of 256 gradations is color converted to the representative value of 11.

  The RGB value converter 304 converts 256 × 256 × 256 = 16,777,216 colors into 11 × 11 × 11 = 1,331 colors.

(4) Specific Example of Model Data Calculation Process by Model Data Calculation Unit 305 Next, a specific example of the model data calculation process by the model data calculation unit 305 will be described. FIG. 7 is a diagram showing a specific example of model data calculation processing.

  As shown in FIG. 7, model data calculation processing is executed for each area image. The example of FIG. 7 indicates that model data calculation processing has been performed for the area image 421 (area name = "area 1"). The model data calculation unit 305 classifies the pixels included in the area image 421 (area name = “area 1”) into 1,331 colors, and counts the number of pixels of each color to obtain model data candidates 700. Generate

  As shown in FIG. 7, the model data candidate 700 includes “color” and “count number” as items of information, and “color” stores 1,331 colors (RGB values), “ The number of pixels of each color included in the area image (area name = "area 1") is stored in the count number.

(5) Specific Example of Model Data Selection Process by Selection Unit 306 Next, a specific example of the model data selection process by the selection unit 306 will be described. FIG. 8 is a diagram showing a specific example of model data selection processing.

  In the example of FIG. 8, the selection unit 306 compares the model data candidates 700, 801, and 802 generated by the model data calculation unit 305 with each other for the region image 421 (region name = "region 1"). It shows. According to the example of FIG. 8, since the model data candidate 700 and the model data candidate 801 have high similarity, the selecting unit 306 selects the model data candidate 700 for the set of model data candidates, and sets the model data as model data. It is registered in the storage unit 123. That is, the model data candidate 801 is not registered in the model storage unit 123.

  On the other hand, since the model data candidate 700 and the model data candidate 802 have a low degree of similarity, the selecting unit 306 registers both of the set of model data candidates in the model storage unit 123 as model data.

  As described above, among the model data candidates generated by the model data calculation unit 305, the model data candidates similar to each other are registered in the model storage unit 123 by selecting and registering one of them. The number of model data is reduced.

<Specific Example of Model Data Registered in Model Storage Unit of Inspection Device>
Next, a specific example of model data registered in the model storage unit 123 of the inspection device 120 will be described. FIG. 9 is a diagram showing a specific example of model data. As shown in FIG. 9, model data is registered in association with an area name and a model name. Model data 911 and 912 are model data of model names = models 1_1 and 1_2, which are used for comparison with an area image of area name = "area 1" among inspection images.

  Similarly, model data 921 and 922 are model data of model names = models 2_1 and 2_2 used for comparison with an area image of area name = "area 2" among inspection images. Similarly, model data 951 and 952 are model data of model name = models 5_1 and 5_2 used for comparison with an area image of area name = "area 5" among inspection images.

<Functional Configuration of Inspection Unit of Inspection Device>
Next, the functional configuration of the inspection unit 122 of the inspection device 120 will be described. FIG. 10 is a diagram showing an example of a functional configuration of the inspection unit of the inspection apparatus. As shown in FIG. 10, the inspection unit 122 includes an image data acquisition unit 1001, an area division unit 1002, a compression unit 1003, an RGB value conversion unit 1004, an inspection data calculation unit 1005, and a determination unit 1006.

  Among the units included in the inspection unit 122, the functions of the image data acquisition unit 1001 to the RGB value conversion unit 1004 are the same as the functions of the image data acquisition unit 301 to the RGB value conversion unit 304 included in the generation unit 121. In order to correspond, explanation is omitted here.

  The inspection data calculation unit 1005 is an example of a calculation unit, and for each area image after color conversion, the pixels included in the area image are classified for each color, and the number of pixels of each color is calculated. Calculate Further, the inspection data calculation unit 1005 notifies the determination unit 1006 of the inspection data of each calculated area.

  The determination unit 1006 is an example of a comparison unit, reads out model data associated with the area to be determined from the model storage unit 123, and compares it with inspection data of the area to be determined, and as a comparison result, the error rate (The details will be described later). Further, the determination unit 1006 is an example of the determination unit, determines the pass / fail of the article to be inspected based on the calculated error rate, and outputs the inspection result.

<Flow of determination processing by determination unit>
Next, the flow of the determination process by the determination unit 1006 will be described. FIG. 11 is a flowchart showing the flow of determination processing. When inspection data of each area is notified from the inspection data calculation unit 1005, the determination unit 1006 starts the flowchart shown in FIG.

  In step S1101, the determination unit 1006 sets 1 in a counter n that counts the number of areas.

  In step S1102, the determination unit 1006 extracts inspection data of the area n from the notified inspection data.

  In step S1103, the determination unit 1006 sets 1 in a counter m that counts the number of model data.

  In step S1104, the determination unit 1006 reads model data n_m of the area n from the model storage unit 123.

  In step S1105, the determination unit 1006 calculates an error rate between the inspection data of the area n and the model data n_m. Specifically, the determination unit 1006 calculates, for each of the 1,331 colors, the difference (absolute value) in the number of pixels between the inspection data of the region n and the model data n_m. In addition, the determination unit 1006 obtains a total value of the calculated difference (absolute value) of the number of pixels, and divides the number by the number of pixels included in the area n (the number of pixels after compression) to calculate an error rate.

  In step S1106, the determination unit 1006 determines whether or not error rates have been calculated for all model data registered in association with the area n. If it is determined in step S1106 that there is model data for which the error rate is not calculated (in the case of No in step S1106), the process proceeds to step S1107.

  In step S1107, the determination unit 1006 increments a counter m that counts the number of model data, and the process returns to step S1104. Thus, the determination unit 1006 can calculate an error rate for the next model data registered in association with the area n.

  On the other hand, if it is determined in step S1006 that the error rates have been calculated for all model data registered in association with the area n (in the case of Yes in step S1106), the process proceeds to step S1108.

  In step S1108, the determination unit 1006 determines whether or not there is model data having an error rate equal to or less than the threshold among the error rates calculated for the inspection data in the area n. If it is determined in step S1108 that model data having an error rate equal to or less than the threshold value is present (in the case of Yes in step S1108), the process proceeds to step S1109.

  In step S1109, the determination unit 1006 determines that the region n is “OK” among the regions included in the article to be inspected.

  On the other hand, when it is determined in step S1108 that there is no model data having an error rate equal to or less than the threshold (No in step S1108), the process proceeds to step S1110.

  In step S1110, the determination unit 1006 determines that the item to be inspected is “failed”. That is, when one of the regions included in the inspection target article is NG, the determination unit 1006 determines that the inspection target article itself is “failed”.

  In step S1111, the determination unit 1006 determines whether the process has been performed for all the areas included in the item to be inspected. If it is determined in step S1111 that there is an area for which processing has not been performed (in the case of No in step S1111), the process proceeds to step S1112.

  In step S1112, the determination unit 1006 increments the counter n for counting the number of areas, and the process returns to step S1102. Thereby, the determination unit 1006 can calculate the error rate for the next area included in the article to be inspected.

  On the other hand, if it is determined in step S1112 that the processing has been performed for all the areas included in the inspection target item (in the case of Yes in step S1111), the process proceeds to step S1113.

  In step S1113, the determination unit 1006 determines that the item to be inspected is “pass”. That is, when all the areas included in the item to be inspected are “OK”, the determination unit 1006 determines that the item to be inspected is “pass”.

  In step S1114, the determination unit 1006 determines whether to end the determination process. If it is determined in step S1114 that the determination process is to be continued (in the case of No in step S1114), the process returns to step S1101. In this case, when the next inspection data is notified, the above processing (steps S1101 to S1113) is executed. On the other hand, when it is determined in step S1114 that the determination process is not to be continued (in the case of Yes in step S1114), the determination process ends.

<Specific Example of Determination Process by Determination Unit>
Next, a specific example of the determination process for the article to be inspected by the determination unit 1006 will be described. FIG. 12 is a diagram illustrating a specific example of the determination process.

  As shown in FIG. 12, when the inspection data 1201 of the area name = “area 1” is notified, the determination unit 1006 causes the model storage unit 123 to register the model registered in association with the area name = “area 1”. Read data sequentially. The example of FIG. 12 indicates that model data 911 (model name = “model data 1_1”) is read from the model storage unit 123.

  Further, as shown in FIG. 12, the determination unit 1006 calculates, for each color, the difference (absolute value) in the number of pixels between the inspection data 1201 and the model data 911, and generates difference data 1202.

  According to the difference data 1202, in the case of color = (0, 0, 0), the difference (absolute value) in the number of pixels between the inspection data 1201 and the model data 911 is “2”. In the case of color = (25, 0, 0), the difference (absolute value) in the number of pixels between the inspection data 1201 and the model data 911 is “4”. Furthermore, in the case of color = (50, 0, 0), the difference (absolute value) in the number of pixels between the inspection data 1201 and the model data 911 is “1”.

  Hereinafter, the determination unit 1006 calculates the difference (absolute value) in the number of pixels between the inspection data 1201 and the model data 911 for all colors, and the total value of the difference (absolute value) in the calculated number of pixels (total difference number Ask for). The example in FIG. 12 indicates that the determination unit 1006 has calculated "90" as the total difference number.

  In addition, the determination unit 1006 further divides the calculated total number of differences by the number of pixels (number of pixels after compression) included in the area name = “area 1” to calculate the error rate, and the area name It is determined whether the "area 1" is "OK".

<Effect of inspection by inspection device>
Next, the effect of the inspection by the inspection device 120 will be described. FIG. 13 is a diagram for explaining the effect of the inspection device. As illustrated in FIG. 13, the generation unit 121 executes compression processing and RGB value conversion processing on the area image 1301 to generate a color-converted area image 1302. In addition, the generation unit 121 registers model data 1303 by executing model data calculation processing and model data selection processing on the color-converted area image 1302.

  On the other hand, the inspection unit 122 performs compression processing and RGB value conversion processing on the area image 1311 to generate a color-converted area image 1312. Further, the inspection unit 122 calculates inspection data 1313 by executing inspection data calculation processing on the color-converted area image 1312.

  As a result, in the inspection unit 122, the determination unit 1006 calculates the difference data 1320, and further calculates “10” as the total difference number. In the example of FIG. 13, since the number of pixels included in the color-converted area image 1302 or 1312 is “132”, the error rate = the total number of differences / the number of pixels = 10/132 = 7.6%. That is, in the inspection unit 122, the determination unit 1006 determines that the area corresponding to the color-converted area image 1312 is “OK”.

  Here, in the area image 1301 and the area image 1311, although the contents being projected are the same, the arrangement of ingredients included in the contents, the size and shape of each ingredient, the color, etc. I do not do it. Therefore, according to the conventional inspection method in which the area image 1301 and the area image 1311 are matched, there is a possibility that the area corresponding to the area image 1311 is determined as “NG”.

  On the other hand, according to the inspection apparatus 120 according to the first embodiment, since the compression process and the RGB value conversion process are performed, the difference in size, shape, color, etc. of the individual ingredients is absorbed. (Refer to the color converted area image 1302 and 1312). On the other hand, if the contents of the image are the same, the color tone of the entire area image matches, and if the difference in the number of pixels classified into each color is calculated, it is determined that the colors match. can do. As described above, according to the inspection apparatus 120 according to the first embodiment, the area corresponding to the area image 1311 can be correctly determined as “OK”.

  That is, according to the inspection apparatus according to the first embodiment, it is possible to improve the inspection accuracy in the appearance inspection of the article (the lunch box).

Second Embodiment
In the first embodiment, the area dividing units 302 and 1002 each divide the inspection image into five areas according to the type of contents, but the number of areas to be divided is limited to five. I will not. Also, the division method is not limited to the type of contents, and may be divided independently of the type of contents (at a predetermined position).

  In the first embodiment, the compression units 303 and 1003 each compress a pixel region of 3 pixels × 3 pixels (9 pixels) into one pixel, but the compression method is not limited to this.

  In the first embodiment, the RGB value conversion units 304 and 1004 respectively convert 25 gradations into one gradation, but the color conversion method by the compression units 303 and 1003 is limited to this. I will not.

  In the first embodiment, the same processing is performed on each area image. However, the processing content may be changed for each area image. For example, the compression method or the color conversion method may be changed for each area image, or the threshold for determining the error rate may be changed for each area image.

  In the first embodiment, the processing is performed on all the area images included in the inspection image as the processing target. However, the processing is performed on the partial area image included in the inspection image as the processing target. It may be performed.

  In the first embodiment, the total difference number is calculated by simply adding the difference in the number of pixels of each color, but the total difference number is calculated by weighted addition of the difference in the number of pixels of each color May be

  In the first embodiment, although the case of the inspection object = the lunchbox has been described, the inspection object is not limited thereto, and the appearance is the same as a product such as bread, vegetables, fruits, and flowers. The present invention is also applicable to other articles, as long as the article is subjected to a pass / fail determination based on the color.

In addition, in the disclosed technology, a form such as that described below can be considered.
(Supplementary Note 1)
Calculation means for classifying the pixels included in a predetermined area according to colors and calculating the number of pixels of each color in the inspection image obtained by photographing the article to be inspected;
A comparison unit that refers to the number of pixels of each color of the pixels included in the area corresponding to the predetermined area in the model image, and compares the calculated number of pixels of each color;
And a determination unit that determines the acceptability of the item to be inspected based on the comparison result.
(Supplementary Note 2)
The determination means
It is characterized by judging the pass / fail of the article to be inspected based on the error rate obtained by dividing the total value of the difference of the number of pixels of each color by the number of pixels included in the predetermined area. The inspection apparatus according to 1.
(Supplementary Note 3)
The determination means
When it is determined that the error rate calculated for the predetermined area is not less than or equal to a predetermined threshold, it is determined that the article to be inspected is rejected.
When it is determined that the error rates calculated for all the regions included in the inspection image are all equal to or less than a predetermined threshold value, it is determined that the article to be inspected is acceptable. Inspection equipment.
(Supplementary Note 4)
The inspection apparatus according to claim 3, further comprising conversion means for converting pixel values of pixels included in the predetermined area into representative values predetermined in a gradation range to which the pixel values belong.
(Supplementary Note 5)
The inspection apparatus according to claim 4, further comprising compression means for compressing the pixels included in the predetermined area into pixels having a pixel value of 1 collectively for every pixel of a predetermined number of pixels.
(Supplementary Note 6)
In an inspection image obtained by photographing an article to be inspected, pixels included in a predetermined area are classified by color, and the number of pixels of each color is calculated.
In the model image, the number of pixels of each color of the pixels included in the area corresponding to the predetermined area is referred to and compared with the calculated number of pixels of each color,
Determine the pass / fail of the item to be inspected based on the comparison result;
An inspection method in which a computer executes processing.
(Appendix 7)
On the computer
In an inspection image obtained by photographing an article to be inspected, pixels included in a predetermined area are classified by color, and the number of pixels of each color is calculated.
In the model image, the number of pixels of each color of the pixels included in the area corresponding to the predetermined area is referred to and compared with the calculated number of pixels of each color,
Determine the pass / fail of the item to be inspected based on the comparison result;
Inspection program to execute the process.

  Note that the present invention is not limited to the configurations shown here, such as combinations with other elements in the configurations and the like described in the above embodiments. These points can be modified without departing from the spirit of the present invention, and can be appropriately determined according to the application form.

100: inspection system 110: imaging device 120: inspection device 121: generation unit 122: inspection unit 301: image data acquisition unit 302: area division unit 303: compression unit 304: RGB value conversion unit 305: model data calculation unit 306: selection Parts 911 to 952: model data 1001: image data acquisition part 1002: area division part 1003: compression part 1004: RGB value conversion part 1005: inspection data calculation part 1006: judgment part

Claims (7)

Calculation means for classifying the pixels included in a predetermined area according to colors and calculating the number of pixels of each color in the inspection image obtained by photographing the article to be inspected;
A comparison unit that refers to the number of pixels of each color of the pixels included in the area corresponding to the predetermined area in the model image, and compares the calculated number of pixels of each color;
And a determination unit that determines the acceptability of the item to be inspected based on the comparison result. The determination means
According to an error rate obtained by dividing the total value of the differences in the number of pixels of each color by the number of pixels included in the predetermined area, the pass / fail of the article to be inspected is determined. An inspection device according to Item 1. The determination means
When it is determined that the error rate calculated for the predetermined area is not less than or equal to a predetermined threshold, it is determined that the article to be inspected is rejected.
When it is determined that the error rates calculated for all the regions included in the inspection image are all equal to or less than a predetermined threshold, it is determined that the article to be inspected is a pass. Inspection device described.   4. The inspection apparatus according to claim 3, further comprising conversion means for converting pixel values of pixels included in the predetermined area into representative values predetermined in a gradation range to which the pixel values belong.   5. The inspection apparatus according to claim 4, further comprising compression means for compressing the pixels included in the predetermined area into pixels having a pixel value of 1 collectively for every pixel of a predetermined number of pixels. In an inspection image obtained by photographing an article to be inspected, pixels included in a predetermined area are classified by color, and the number of pixels of each color is calculated.
In the model image, the number of pixels of each color of the pixels included in the area corresponding to the predetermined area is referred to and compared with the calculated number of pixels of each color,
Determine the pass / fail of the item to be inspected based on the comparison result;
An inspection method in which a computer executes processing. On the computer
In an inspection image obtained by photographing an article to be inspected, pixels included in a predetermined area are classified by color, and the number of pixels of each color is calculated.
In the model image, the number of pixels of each color of the pixels included in the area corresponding to the predetermined area is referred to and compared with the calculated number of pixels of each color,
Determine the pass / fail of the item to be inspected based on the comparison result;
Inspection program to execute the process.