Identification system, identification method, and program

Abstract

To prevent reduction in accuracy when identification is performed by using an image of a print with a small contrast between the color of ink used for printing of an object for which identification is performed and the color of a base located in an area of the print where the object is printed.SOLUTION: An identification system comprises: an image acquisition unit that acquires an object image in a print obtained by picking up an image of an object for which identification is performed; a preprocessing execution determination unit that, on the basis of the distribution of the pixel values of the pixels constituting the object image, determines whether to perform preprocessing on the object image; and a preprocessing unit that, when performing the preprocessing on the object image, adds or subtracts a uniform offset value to or from pixel values of a group of pixels, of the pixels constituting the object image, whose pixel values are included in a specific range, and thereby changes the pixel values of the object image.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an individual identification system, an individual identification method, and a program. [Background technology]

[0002] There is known an individual identification system that uses feature points due to printing variations (for example, Patent Document 1). In such an individual identification system, individual identification is performed by extracting feature points such as blurs and defects that occurred during printing and differences due to printing methods from an image of the printed matter. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2019-139640 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, it is difficult to extract feature points when the contrast of an image is low, such as when the color of the ink used to print the object to be identified is similar to the color of the background in the area of ​​the printed matter where the object is to be printed. On the other hand, when an image with low contrast is binarized to increase the contrast, there is a problem in that feature points such as blurring are lost, reducing the accuracy of individual identification.

[0005] The present invention has been made in light of the above-mentioned problems, and aims to provide an individual identification system, an individual identification method, and a program that can suppress a decrease in accuracy when performing individual identification using an image of a printed matter in which there is low contrast between the color of the ink used to print the object to be individually identified and the color of the background in the area of ​​the printed matter where the object is printed. [Means for solving the problem]

[0006] The individual identification system of the present invention comprises an image acquisition unit that acquires an object image of an object that is to be individually identified among printed matter, a preprocessing execution determination unit that determines whether or not to perform preprocessing on the object image based on the distribution of pixel values ​​of each of the pixels that make up the object image, and a preprocessing unit that, when preprocessing is to be performed on the object image, changes the pixel values ​​of the object image by adding or subtracting a uniform offset value to each pixel value of a group of pixels that make up the object image and whose pixel values ​​fall within a specific range.

[0007] The individual identification method of the present invention is an individual identification method performed by a computer that performs individual identification, which obtains an object image in which an object that is to be individually identified among printed matter is captured, and determines whether or not to perform preprocessing on the object image based on the distribution of pixel values ​​of each of the pixels that make up the object image, and if preprocessing is to be performed on the object image, changes the pixel values ​​of the object image by adding or subtracting a uniform offset value to each pixel value of a group of pixels that make up the object image and whose pixel values ​​fall within a specific range.

[0008] The program of the present invention causes a computer operating an individual identification system to acquire an object image in which an object that is to be individually identified among printed matter is captured, and determines whether or not to perform preprocessing on the object image based on the distribution of pixel values ​​of each of the pixels that make up the object image, and if preprocessing is to be performed on the object image, changes the pixel values ​​of the object image by adding or subtracting a uniform offset value to each pixel value of a group of pixels that make up the object image and whose pixel values ​​fall within a specific range. Effect of the Invention

[0009] According to the present invention, it is possible to suppress a decrease in accuracy when performing individual identification using an image of a printed matter in which there is little contrast between the color of the ink used to print the object to be identified and the color of the background in the area of ​​the printed matter where the object is printed. [Brief description of the drawings]

[0010] [Figure 1] 1 is a block diagram showing an example of the configuration of an individual identification system 1 according to an embodiment. [Figure 2A] FIG. 2 is a diagram showing an example of an object image G according to the embodiment. [Figure 2B] FIG. 13 is a diagram showing an example of an object image G converted into a grayscale image according to the embodiment. [Figure 2C] FIG. 3 is a diagram showing an example of a distribution based on the object image G of FIG. 2B. [Figure 3A] FIG. 2 is a diagram showing an example of an object image G that has been preprocessed according to the embodiment. [Figure 3B] FIG. 3B is a diagram showing an example of a distribution based on the object image G of FIG. 3A. [Figure 4] 4 is a diagram showing an example of information stored in an image information storage unit 17 according to the embodiment. FIG. [Diagram 5] 4 is a diagram showing an example of information stored in an image information storage unit 17 according to the embodiment. FIG. [Figure 6] 2 is a diagram for explaining a process performed by the information processing device 10 of the embodiment. FIG. [Figure 7] 4 is a flowchart showing a flow of processing performed by the information processing apparatus 10 of the embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, an individual identification system 1 and an information processing device 10 according to an embodiment will be described with reference to the drawings. In the following description, it is assumed that an arbitrary object having characteristics such as smudges or chips that may occur during printing is attached to an individual that is the subject of individual identification, and individual identification is performed based on the characteristics of the object attached to the individual. The object may be attached directly to the individual, for example, by being directly printed on the individual, or the object may be attached indirectly to the individual, for example, by wrapping the individual in a packaging material on which the object is printed. The target also includes a printed area in which some symbol or the like is printed, and a blank area (quiet zone) provided around the printed area. In the following, a case will be described in which the target is a printed area of ​​a two-dimensional code attached to an individual, and a quiet zone provided around the two-dimensional code. However, the target is not limited to this. The target can be any printed area in which individual differences such as smudges or chipping may occur during printing, such as printed characters, symbols, graphics such as logos, one-dimensional codes, JAN (Japanese Article Number) codes, images, and combinations of these, and the quiet zone provided around the printed area.

[0012] (About Individual Identification System 1) The individual identification system 1 will be described with reference to Fig. 1. Fig. 1 is a block diagram showing an example of the configuration of the individual identification system 1 according to an embodiment. As shown in Fig. 1, the individual identification system 1 includes, for example, an information processing device 10, an imaging device 20, and a display device 30. The information processing device 10 and the imaging device 20, and the information processing device 10 and the display device 30 are communicatively connected via a communication network such as the Internet, or via a short-range wireless communication method such as Bluetooth (registered trademark) or infrared communication, or via a wire such as a USB cable.

[0013] The imaging device 20 is a camera that captures an image of a two-dimensional code (hereinafter, referred to as an object) printed on a printed matter as a target for individual identification. The imaging device 20 captures an image of the object and transmits image information of the captured image of the object (hereinafter, referred to as an object image G) to the information processing device 10.

[0014] The information processing device 10 is a computer that performs individual identification. The information processing device 10 receives image information of an object image G from the imaging device 20, and performs individual identification of the object using the received image information. The information processing device 10 transmits the identification result to the display device 30.

[0015] The display device 30 is a display that displays images, etc. The display device 30 receives the identification result obtained by individually identifying the object from the information processing device 10, and displays the identification result. Although the case where the identification result is displayed on the display device 30 has been described here as an example, the present invention is not limited to this. In the individual identification system 1, the identification result may be output by a sound such as an alarm sound or a voice. In this case, a sound output device (speaker) is provided in place of the display device 30 or in addition to the display device 30. The speaker outputs a sound corresponding to the identification result received from the information processing device 10. Alternatively, the identification result may be printed out in the individual identification system 1. In this case, a printer is provided in place of or in addition to the display device 30. The printer prints out and outputs the identification result received from the information processing device 10.

[0016] (Regarding information processing device 10) As shown in FIG. 1, the information processing device 10 includes, for example, an image acquisition unit 11, a grayscale conversion unit 12, a preprocessing execution determination unit 13, a preprocessing unit 14, an individual identification unit 15, an output unit 16, an image information storage unit 17, and a feature point storage unit 18.

[0017] The image acquisition unit 11 acquires image information of the object image G. When an object is imaged by the imaging device 20, the image acquisition unit 11 acquires image information of the object image G of the object from the imaging device 20. The image information here is information in which a pixel value indicating a color (e.g., RGB value) is associated with each pixel. The image acquisition unit 11 outputs the image information of the acquired image to the grayscale conversion unit 12.

[0018] The grayscale conversion unit 12 generates an object image G in which pixel values ​​(e.g., RGB values) are converted to grayscale. The object image G converted to grayscale is an image in which pixel values ​​range from 0 to 255, for example, with a pixel value of 0 indicating the darkest black color and a pixel value of 255 indicating the brightest white color. Any method may be used to convert the object image G to grayscale. Methods for converting a color image to a grayscale image are well known, and therefore description thereof will be omitted. The grayscale conversion unit 12 outputs the object image G converted to grayscale to the preprocessing execution determination unit 13.

[0019] The pre-processing execution determination unit 13 determines whether or not to execute pre-processing. The pre-processing here is a process performed before individual identification of the object image G, and is a process for suppressing a decrease in accuracy of the individual identification. In this embodiment, in particular, a process for suppressing a decrease in accuracy that may occur when the contrast of the object image G is low is executed as the pre-processing. The contrast here is the difference between the color of the printed matter that is the background of the printed matter and the color of the printed object.

[0020] The preprocessing execution determination unit 13 determines whether the object image G has a contrast sufficient for accurate individual identification (hereinafter referred to as distinguishable contrast). If the object image G has distinguishable contrast, the preprocessing execution determination unit 13 determines not to execute preprocessing on the object image G. On the other hand, if the object image G does not have distinguishable contrast, the preprocessing execution determination unit 13 determines to execute preprocessing on the object image G. The preprocessing execution determination unit 13 outputs the determination result to the preprocessing unit 14.

[0021] The pre-processing execution determination unit 13 determines whether the object image G has a distinguishable contrast based on the distribution of pixels that constitute the object image G converted into grayscale.

[0022] Here, a method for the pre-processing execution determination unit 13 to determine whether or not the object image G has a distinguishable contrast will be described with reference to Fig. 2 (Figs. 2A to 2C). Fig. 2 is a diagram for explaining the process performed by the information processing device 10 of the embodiment.

[0023] An example of an object image G1 is shown in Fig. 2A. As shown in Fig. 2A, the object image G1 is an image captured of an object printed in ink with a low brightness, such as light blue, on a light gray background. 2B shows an example of object image G2, which is an image in which the pixel values ​​of object image G1 are converted to grayscale. 2C is a schematic diagram showing a histogram indicating the distribution of pixels constituting the object image G2. In the histogram in FIG. 2C, the horizontal axis indicates pixel value, and the vertical axis indicates frequency. 2C, in a histogram of pixels constituting the object image G2, for example, there are pixel groups Gr1 and Gr2 that are distributed with two peak points P1 and P2 with high frequencies (frequency) as apexes. The first pixel group Gr1 is a pixel group made up of pixels that belong to a specific pixel value range E1 around the pixel value D1, with the peak point P1 (pixel value D1, frequency N1) as its apex. The second pixel group Gr2 is a pixel group made up of pixels that belong to a specific pixel value range E2 around the pixel value D2, with the peak point P2 (pixel value D2, frequency N2) as its apex. When converting object image G to grayscale, if pixel value 0 indicates the darkest black color and pixel value 255 indicates the brightest white color, pixel value D1 indicates a color closer to black than pixel value D2. In this case, pixel group Gr1 indicates the distribution of pixels that make up the part of object image G2 where the object is printed, and pixel group Gr2 indicates the distribution of pixels that make up the background part of object image G2.

[0024] The pre-processing execution determination unit 13 calculates the distribution of pixels that form the object image G. The pre-processing execution determination unit 13 determines whether the calculated distribution satisfies the following two conditions. First condition: There are two pixel groups Gr that are distributed with a peak as their apex. Second condition: The difference between the representative pixel values ​​of the two pixel groups Gr is equal to or greater than a first threshold value (e.g., 80).

[0025] The representative pixel value in the second condition is a pixel value that represents each pixel value of the pixels that make up the pixel group Gr. The representative pixel value is calculated, for example, by applying a statistical process to each pixel value of the pixels that make up the pixel group. The representative pixel value is, for example, a simple average value, a mode, a maximum value, a median, or the like, of each pixel value of the pixels that make up the pixel group. In the example of FIG. 2C, the representative pixel value of pixel group Gr1 is pixel value D1. The representative pixel value of pixel group Gr2 is pixel value D2. The difference between the representative pixel values ​​of the two pixel groups Gr is |D2-D1|. This difference is an example of an index that indicates the magnitude of contrast between the two main colors (pixel values ​​converted to grayscale) in the object.

[0026] When both the first condition and the second condition are satisfied in the distribution of pixels constituting the object image G, the preprocessing execution determination unit 13 determines that the object image G has a distinguishable contrast. In this case, the preprocessing execution determination unit 13 determines not to execute preprocessing on the object image G.

[0027] On the other hand, when the distribution of pixels constituting the object image G satisfies the first condition but does not satisfy the second condition, the preprocessing execution determination unit 13 determines that the object image G does not have distinguishable contrast. In this case, the preprocessing execution determination unit 13 determines to execute preprocessing on the object image G.

[0028] It should be noted that, in the case where the distribution of pixels constituting the object image G does not satisfy the first condition, the pre-processing execution determination unit 13 determines that individual identification of the object image G is not to be performed. Here, cases in which the first condition is not satisfied include a first case in which there is only one pixel group Gr whose distribution has a peak as its apex in the distribution of the pixels that make up the object image G, and a second case in which there are three or more pixel groups Gr whose distribution has a peak as its apex in the distribution of the pixels that make up the object image G. The first case indicates that it is difficult to distinguish between the background color and the ink color in the object image G. Such an object image G is excluded from the target of individual identification in this embodiment. In the second case, the background in the object image G is made up of multiple colors, and / or the object is printed using ink of multiple colors. Such object images G are excluded from the individual identification target in this embodiment.

[0029] The preprocessing execution determination unit 13 stores image information of the object image G in the image information storage unit 17. The preprocessing execution determination unit 13 also stores information indicating the distribution of pixels constituting the object image G in the image information storage unit 17. The preprocessing execution determination unit 13 also stores in the image information storage unit 17 the result of the determination as to whether or not to execute preprocessing on the object image G.

[0030] Returning to the explanation of Fig. 1, the preprocessing unit 14 executes preprocessing. The preprocessing unit 14 executes preprocessing on the object image G for which the preprocessing execution determination unit 13 has determined that preprocessing should be executed, that is, the object image G that does not satisfy the second condition.

[0031] The pre-processing unit 14 performs pre-processing so that the object image G after pre-processing satisfies the second condition. The pre-processing unit 14 adds or subtracts a uniform offset value X to each pixel value of pixels constituting at least one of the two pixel groups Gr so that the difference between the representative pixel values ​​of the two pixel groups Gr extracted under the first condition is equal to or greater than a first threshold value (e.g., 80). The pre-processing unit 14 performs a process of changing each pixel value by adding or subtracting such a uniform offset value as pre-processing. By adding or subtracting a uniform offset value X, it is possible to prevent the disappearance of feature points used for individual identification, such as blurring. By changing the pixel values ​​of each of the pixels constituting the pixel group Gr, the difference can be made equal to or greater than a first threshold value (e.g., 80). In other words, the pre-processing unit 14 can increase the contrast between the object and the background without causing the disappearance of feature points. Therefore, it is possible to suppress a decrease in accuracy when performing individual identification using an image of an object with low contrast between the ink (object) and the background.

[0032] Here, a method in which the preprocessing unit 14 performs preprocessing on the object image G will be described with reference to Fig. 3 (Figs. 3A and 3B). Fig. 3 is a diagram for explaining the processing performed by the information processing device 10 of the embodiment.

[0033] 3A shows an example of an object image G3. The object image G3 is an image obtained by performing preprocessing on an object image G (here, the object image G2 in FIG. 2B) that does not satisfy the second condition. The pre-processing unit 14 executes a pre-processing process of subtracting a uniform offset value X from each pixel value of a pixel group Gr1 that is distributed around the peak point P1 in Fig. 2C among the pixels that constitute the object image G2 in Fig. 2B. By executing such pre-processing, each pixel value of the pixels that constitute the pixel group Gr1 becomes a value having the uniform offset value X subtracted therefrom.

[0034] 3B is a schematic diagram showing a histogram indicating the distribution of pixels constituting the object image G3. In the histogram in FIG. 3B, the horizontal axis indicates pixel value, and the vertical axis indicates frequency. As shown in Fig. 3B, in the histogram of the pixels constituting the object image G3, there are pixel groups Gr3 and Gr2 that are distributed with two peak points P3 and P2 with high frequencies (frequency) as apexes, just like in the object image G2. The first pixel group Gr3 is a pixel group composed of pixels that belong to a specific pixel value range E3 around the pixel value D3 (pixel value D3, frequency N1) as apex. The second pixel group Gr2 is equivalent to the pixel group Gr2 in Fig. 2C.

[0035] The pixels constituting pixel group Gr3 are obtained by changing the pixel values ​​of the pixels constituting pixel group Gr1 in object image G2 through pre-processing. In other words, as a result of changing the pixel values ​​of the pixels constituting pixel group Gr1, pixel group Gr3 is formed, which is distributed with peak point P3 as its apex. Representative pixel value D3 in pixel group Gr3 is the representative pixel value D1 minus a uniform offset value X. In other words, "D3=D1-X".

[0036] The difference between the representative pixel values ​​of the two pixel groups Gr3 and Gr2 in the object image G3 is |D2-D3|, i.e. |D2-D1+X|, which is a larger value than the difference |D2-D1| in the object image G2. For this reason, the preprocessing unit 14 appropriately sets the offset value X and performs preprocessing, so that the difference between the representative pixel values ​​of the two pixel groups Gr3 and Gr2 in the object image G3 is equal to or larger than a first threshold value (e.g., 80), and the preprocessed object image G satisfies the second condition.

[0037] Returning to the explanation of Fig. 1, the individual identification unit 15 performs individual identification based on feature points of the object image G. The feature points here are features of the object such as smudges or chips caused by printing.

[0038] As a prerequisite for performing individual identification, it is assumed that information indicating feature points extracted from a reference image is stored in the feature point storage unit 18. The reference image here is an image that is used as a reference when performing individual identification.

[0039] For example, before the shipment of the product, an object attached to the product is captured as a reference image. Then, feature points are extracted from the reference image. Any method may be used to extract feature points from the reference image. For example, it is possible to extract feature points from the reference image using image processing that extracts contours from an image, or machine learning such as deep learning. The feature points extracted from the reference image are stored in the feature point storage unit 18 as quantified feature point data.

[0040] Thereafter, the product is shipped, and the object attached to the product is captured by the user who has purchased the product, and the image (object image G) is transmitted to the information processing device 10, thereby entering a campaign, etc. In such a case, the individual identification unit 15 performs individual identification of the object image G. If it is found that the object image G is an image of the same individual (object) as the object whose reference image was captured before shipping, it can be determined, for example, that the product purchased by the user was shipped via a regular shipping route and is eligible to enter a campaign, etc. Alternatively, as an application for a campaign or the like for a product, an image of an object attached to the product is captured by a smartphone or the like of a user who has purchased the product as an application image (referred to as a first application image) and transmitted to the information processing device 10. Feature points of the object extracted from the first application image are stored in the feature point storage unit 18 as registered feature point data. If it is subsequently found that an application image used for application (referred to as a second application image) has the same feature points as the registered feature point data, it is possible to regard the application as having already been submitted with respect to the first application image, and reject the application with respect to the second application image. Of course, the individual identification may be used for purposes other than applying for a campaign.

[0041] The individual identification unit 15 performs individual identification of the object image G by referring to information (reference feature point data) stored in the feature point storage unit 18. The individual identification unit 15 extracts feature points from the object image G, and refers to the feature point storage unit 18 based on information indicating the extracted feature points (feature point data of the object). The individual identification unit 15 calculates the correlation, for example, the similarity, between each of the reference feature point data and the feature point data of the object. Any method can be adopted as a method for calculating the similarity. For example, the individual identification unit 15 can calculate the similarity of the distance between the coordinate value of the reference feature point data and the coordinate value of the feature point data of the object in the feature amount space. In this case, the smaller the distance, the more similar the data are, and the larger the distance, the less similar the data are. The feature amount space here is a space expressed by a feature vector according to feature point data, and for example, when the feature point data is data expressed in three dimensions consisting of a first dimension corresponding to a feature vector indicating the degree of print smearing, a second dimension corresponding to a feature vector indicating the degree of print loss, and a third dimension corresponding to a feature vector indicating the printing method, the feature amount space becomes a three-dimensional space. Feature point data according to coordinate points corresponding to each feature amount (degree of smearing, degree of loss, printing method) can be mapped into the feature amount space. It should be noted that the feature vectors that make up the feature space are not limited to feature vectors indicating the degree of print blur, feature vectors indicating the degree of print loss, and feature vectors indicating the printing method, or combinations of these, and it goes without saying that other feature vectors may also be applied.

[0042] When it is shown that the similarity between the feature point data of the object and the feature point data of the reference destination is equal to or greater than a predetermined threshold value, the individual identification unit 15 determines that the object imaged in the object image G and the object imaged in the reference image are the same individual. On the other hand, when it is shown that the similarity between the feature point data of the object and the feature point data of the reference destination is not equal to or greater than a predetermined threshold value, it determines that the object imaged in the object image G and the object imaged in the reference image are not the same individual.

[0043] The output unit 16 outputs the identification result by the individual identification unit 15. The output unit 16 outputs the identification result to the display device 30. As a result, the display device 30 displays the identification result. In addition, in the individual identification system 1, when the identification result is output by a sound such as an alarm sound or a voice, the output unit 16 outputs, to the speaker, information indicating a sound corresponding to the identification result by the individual identification unit 15. The speaker displays the sound corresponding to the identification result. Alternatively, in the individual identification system 1, when the identification result is printed out, the output unit 16 outputs, to the printer, information for printing the identification result by the individual identification unit 15. The printer prints and outputs the identification result.

[0044] The image information storage unit 17 stores information about the object image G. The feature point storage unit 18 stores information about feature points extracted from the reference image.

[0045] The storage unit (including the image information storage unit 17 and the feature point storage unit 18) included in the information processing device 10 is configured by a storage medium such as a hard disk drive (HDD), a flash memory, an electrically erasable programmable read only memory (EEPROM), a random access read / write memory (RAM), a read only memory (ROM), or a combination of these. The storage unit included in the information processing device 10 stores programs for executing various processes for realizing the functions of the information processing device 10, and temporary data used when performing various processes.

[0046] The functional units (including the image acquisition unit 11, the grayscale conversion unit 12, the preprocessing execution determination unit 13, the preprocessing unit 14, the individual identification unit 15, and the output unit 16) of the information processing device 10 are realized by having a CPU (Central Processing Unit) and / or a GPU (Graphics Processing Unit) that the information processing device 10 has as hardware execute a program stored in a memory unit that the information processing device 10 has.

[0047] 4 and 5 are diagrams showing examples of information stored in the image information storage unit 17 of the embodiment. The image information storage unit 17 stores information about the object image G for each object image G. The information about the object image G includes, for example, image information about the object image G, information indicating the distribution of pixels constituting the object image G, and a determination result as to whether or not to perform preprocessing on the object image G. The information about the object image G is stored in the image information storage unit 17 by, for example, the preprocessing execution determination unit 13.

[0048] Fig. 4 shows an example of information stored in the image information storage unit 17 when it is determined that "preprocessing is performed" as a result of the determination as to whether or not to perform preprocessing on the object image G. Fig. 5 shows an example of information stored in the image information storage unit 17 when it is determined that "preprocessing is not performed" as a result of the determination as to whether or not to perform preprocessing on the object image G.

[0049] 4 and 5, information corresponding to each of the items of image information, statistics, and judgment results is stored in the image information storage unit 17. The image information is image information of an object image G, and is, for example, information indicating pixel coordinates and pixel values ​​of each pixel group constituting the object image G. The statistics are information indicating the distribution of pixels constituting the object image G. As the statistics, for example, information indicating the frequency and representative pixel value of each of the two peaks, and the difference between the representative pixel values ​​of the two peaks is stored. The determination result is information indicating whether or not to perform preprocessing on the object image G. For example, if the difference between the representative pixel values ​​of the two peaks is equal to or greater than a first threshold value (e.g., 80), it is determined that preprocessing is not to be performed. On the other hand, if the difference between the representative pixel values ​​is less than the first threshold value (e.g., 80), it is determined that preprocessing is to be performed. In the example of FIG. 4, the difference between the representative pixel values ​​is "26", which is less than the first threshold value (for example, 80), so "Perform preprocessing (needed)" is displayed, which indicates that preprocessing should be performed. In the example of FIG. 5, the difference between the representative pixel values ​​is "92", which is equal to or greater than the first threshold value (for example, 80), and therefore "Perform preprocessing (not required)" is displayed, which means that preprocessing is not performed.

[0050] Here, the process performed by the pre-processing execution determination unit 13 will be further described with reference to Fig. 6. Fig. 6 is a diagram for explaining the process performed by the information processing device 10 of the embodiment.

[0051] Fig. 6 shows an example of the distribution of pixel values ​​of the pixels that make up the object image G. In the histogram of Fig. 6, the horizontal axis indicates pixel value, and the vertical axis indicates frequency. As shown in FIG. 6, in an actual object image G, it may be difficult to separate into two pixel groups Gr with a peak at the apex as in FIG. 2C or FIG. 3B. In the example of FIG. 6, there are five peaks corresponding to pixel values ​​D10 to D14. For example, in the example of this figure, it is desirable to extract a pixel group included in a range E10 as a pixel group corresponding to the object, and to extract a pixel group included in a range E11 as a pixel group corresponding to the background. However, if two peaks are mechanically extracted in descending order of frequency in a distribution such as that of FIG. 6, there is a possibility that a pixel group Gr in an incorrect range will be extracted. If preprocessing is performed based on a pixel group Gr that has been erroneously extracted, it becomes difficult to perform individual identification with high accuracy.

[0052] As a countermeasure, in this embodiment, when there are three or more peaks in the distribution of pixel values ​​of pixels constituting the object image G, if the difference between the representative pixel values ​​of adjacent pixel groups among the pixel groups having each peak as an apex is less than a second threshold value (e.g., 40), the two adjacent pixel groups are regarded as one pixel group. That is, even if there are three or more peaks and there are three or more adjacent pixel groups among the pixel groups having each peak as an apex, the two adjacent pixel groups are combined into one pixel group if the distance between the two adjacent pixel groups is close, that is, if the difference between the representative pixel values ​​is less than a second threshold value (e.g., 40). For example, in the example of FIG. 6, of the five pixel groups having five peaks corresponding to pixel values ​​D10 to D14 as apexes extracted from the distribution, four pixel groups having four peaks corresponding to pixel values ​​D11 to D14 as apexes can be combined into one pixel group. This makes it possible to classify a pixel group into two groups, one estimated to correspond to the object and the other estimated to correspond to the background, even if three or more peaks are present.

[0053] FIG. 7 is a flowchart showing a flow of processing performed by the information processing device 10 of the embodiment. First, the preprocessing execution determination unit 13 of the information processing device 10 generates a histogram showing the distribution of pixel values ​​of each pixel constituting the object image G converted to grayscale (step S10). Next, the preprocessing execution determination unit 13 classifies the pixel groups distributed with the peak as the apex (step S11). First, the preprocessing execution determination unit 13 extracts a peak in the histogram. The preprocessing execution determination unit 13 detects a pixel value having a frequency higher than the frequency of an adjacent pixel value as a peak. That is, when there is a point (pixel value D, frequency N) on the histogram and the frequencies of the adjacent pixel value (D-1) and pixel value (D+1) are both smaller than the frequency N, the point (pixel value D, frequency N) in the distribution is extracted as a peak. Then, the preprocessing execution determination unit 13 divides the histogram into pixel groups distributed with the extracted peak as the apex. When there are two extracted peaks, the pre-processing execution determination unit 13 divides the histogram into two pixel groups using a pixel value (for example, (first pixel value+second pixel value) / 2) between a pixel value (first pixel value) corresponding to one peak (first peak) and a pixel value (second pixel value) of the other peak (second peak) as a boundary. When there are three or more extracted peaks, the pre-processing execution determination unit 13 can also divide the histogram into three or more pixel groups using a similar method.

[0054] The pre-processing execution determination unit 13 determines how many pixel groups the histogram has been divided into (step S12). If the histogram has been divided into two pixel groups, the pre-processing execution determination unit 13 executes the process shown in step S13. If the histogram has been divided into three or more pixel groups, the pre-processing execution determination unit 13 executes the process shown in step S16. If the histogram has been divided into one pixel group (i.e., not divided), the pre-processing execution determination unit 13 executes the process shown in step S22, determines that individual identification will not be performed, and ends the process.

[0055] When the histogram is divided into two pixel groups, the preprocessing execution determination unit 13 determines whether or not to perform preprocessing before performing individual identification (step S13). When the difference between the representative pixel values ​​of the two pixel groups is less than a first threshold value (e.g., 80), the preprocessing execution determination unit 13 determines that preprocessing is to be performed before performing individual identification. When it is determined that preprocessing is to be performed, the information processing device 10 executes the process shown in step S14. On the other hand, when the difference between the representative pixel values ​​of the two pixel groups is equal to or greater than a first threshold value (e.g., 80), the preprocessing execution determination unit 13 determines that preprocessing is not to be performed. When it is determined that preprocessing is not to be performed, the information processing device 10 executes the process shown in step S15.

[0056] The pre-processing unit 14 of the information processing device 10 performs pre-processing (step S14). The pre-processing unit 14 adds or subtracts a uniform offset value X to the pixel values ​​of the pixels belonging to at least one of the pixel groups, thereby changing the pixel values ​​of the pixels belonging to one of the pixel groups, so that the difference (the difference between the representative pixel values ​​of the two pixel groups) becomes equal to or greater than a first threshold value (e.g., 80). Alternatively, the pre-processing unit 14 may change the pixel values ​​of both pixel groups by subtracting a uniform offset value X1 from the pixel values ​​of the pixels belonging to one of the pixel groups and adding a uniform offset value X2 to the pixel values ​​of the pixels belonging to the other pixel group, so that the difference becomes equal to or greater than a first threshold value (e.g., 80).

[0057] The individual identification unit 15 of the information processing device 10 performs individual identification (step S15). For the object image G that has been preprocessed by the preprocessing unit 14, the individual identification unit 15 performs individual identification using the object image G that has been subjected to the preprocessing. For example, when generating a histogram of the object image G before preprocessing (the object image G converted to grayscale), the pixel values ​​constituting the histogram are associated with position coordinates (pixel coordinates) in the object image G. By mapping according to the position coordinates (pixel coordinates) associated with each of the pixel values ​​to be histogrammed after preprocessing, the object image G after preprocessing can be generated. On the other hand, for object images G that have not been preprocessed by the preprocessing unit 14, individual identification is performed using the object images G that have not been preprocessed, in this case the object images G that have been converted to grayscale by the grayscale conversion unit 12. After performing the individual identification, the individual identification unit 15 ends the process.

[0058] In step S12, when the histogram is divided into three or more pixel groups, the preprocessing execution determination unit 13 of the information processing device 10 considers whether it is possible to divide the histogram into two pixel groups. The preprocessing execution determination unit 13 extracts two adjacent pixel groups from the three or more pixel groups (step S16). The preprocessing execution determination unit 13 determines whether the difference between the representative pixel values ​​of the two pixel groups extracted in step S16 is less than a second threshold value (for example, 40) (step S17). If the difference is less than the second threshold value (for example, 40), the preprocessing execution determination unit 13 regards the two pixel groups as one pixel group (step S18). On the other hand, if the difference is equal to or greater than the second threshold value (for example, 40), the preprocessing execution determination unit 13 does not regard the two pixel groups as one pixel group. The preprocessing execution determination unit 13 determines whether or not there are still any pixel groups (two adjacent pixel groups) that have not been extracted in step S16 (step S19), and if there are still any pixel groups that have not been extracted, the process returns to step S16. If there are no more pixel groups that have not been extracted, the preprocessing execution determination unit 13 determines whether or not the number of pixel groups has decreased (step S20). If there are two pixel groups that are considered to be one pixel group in step S18, the preprocessing execution determination unit 13 determines that the number of pixel groups has decreased. On the other hand, if there are not two pixel groups that are considered to be one pixel group in step S18, the preprocessing execution determination unit 13 determines that the number of pixel groups has not decreased. If the number of pixel groups has decreased, the preprocessing execution determination unit 13 returns to step S12. If the number of pixel groups has not decreased, the preprocessing execution determination unit 13 determines that individual identification is not to be performed (step S21), and ends the process.

[0059] As described above, the individual identification system 1 of the embodiment includes the image acquisition unit 11, the preprocessing execution determination unit 13, and the preprocessing unit 14. The image acquisition unit 11 acquires an object image G. The object image G is an image obtained by capturing a printed matter (object) that is the object of individual identification. The preprocessing execution determination unit 13 determines whether or not to perform preprocessing on the object image G based on the distribution (e.g., histogram) of pixel values ​​of each pixel constituting the object image G. When performing preprocessing on the object image G, the preprocessing unit 14 changes the pixel values ​​of the object image G by adding or subtracting a uniform offset value X to each pixel value of a pixel group Gr whose pixel values ​​are included in a specific range among the pixels constituting the object image G. Thereby, in the individual identification system 1 of the embodiment, the pixel values ​​of pixels having pixel values ​​in a certain range can be subtracted or added uniformly. For example, when a pixel value in a certain range is a pixel value corresponding to the color of ink (pixel value converted to grayscale), the pixel value of the pixel corresponding to the object can be subtracted. Alternatively, when a pixel value in a certain range corresponds to the color of the background (pixel value converted to grayscale), the pixel value of the pixel corresponding to the background can be added. This makes it possible to increase the color contrast in the printed matter. Moreover, since a uniform offset value X is added or subtracted, even if the pixel value of the pixel corresponding to the object is changed, feature points used for individual identification, such as blurring, are not lost. Therefore, even if the contrast of the printed matter is low, it is possible to increase the contrast without losing the feature points. In other words, it is possible to suppress a decrease in accuracy when performing individual identification using an image of a printed matter in which the contrast between the color of the ink used to print the object and the color of the background in the area where the object is printed on the printed matter is low.

[0060] Furthermore, in the individual identification system 1 of the embodiment, the preprocessing execution determination unit 13 determines to perform preprocessing on the object image G when the pixels constituting the object image G are divided into two pixel groups according to the frequency of their distribution, and the difference between the representative pixel values ​​of the two divided pixel groups is less than a first threshold value (e.g., 80). This allows the individual identification system 1 of the embodiment to perform preprocessing when the contrast is low. That is, preprocessing can be performed on a printed matter with low contrast, and it is possible to suppress a decrease in accuracy caused by low contrast.

[0061] Furthermore, in the individual identification system 1 of the embodiment, when the pixels constituting the object image G are divided into three or more pixel groups according to the frequency of the distribution, the preprocessing execution determination unit 13 regards two adjacent pixel groups, in which the difference between the representative pixel values ​​of each pixel group is less than the second threshold value, as one pixel group. As a result, in the individual identification system 1 of the embodiment, even when the distribution is divided into three or more pixel groups, the two pixel groups can be combined into one pixel group according to the closeness of the representative pixel values ​​in each pixel group, and the pixel groups can be separated into pixel groups that are highly likely to correspond to the color of the ink used to print the object and the color of the background in the area where the object is printed in the printed matter.

[0062] Furthermore, in the individual identification system 1 of the embodiment, the preprocessing unit 14 changes the pixel values ​​of the object image G so that the difference between the representative pixel values ​​of the two pixel groups is equal to or greater than the first threshold value. This allows the individual identification system 1 of the embodiment to perform preprocessing so as to increase the contrast of the printed matter.

[0063] Furthermore, in the individual identification system 1 of the embodiment, the preprocessing unit 14 changes the pixel values ​​of the object image G so that the difference between the representative pixel values ​​of the two pixel groups is equal to or greater than the first threshold value by adding or subtracting a uniform offset value to each pixel value of the pixels included in at least one of the two pixel groups. This makes it possible for the individual identification system 1 of the embodiment to increase the contrast between the two main colors (pixel values ​​converted to grayscale) that make up the printed matter, and to prevent a decrease in the identification accuracy of the individual identification.

[0064] In the above-described embodiment, a case has been described in which the image obtained by converting the object image G to grayscale is used to determine whether or not to perform preprocessing, and when performing preprocessing, the image obtained by converting the object image G to grayscale is used for preprocessing. However, the present invention is not limited to this. It may be possible to determine whether or not to perform preprocessing based on the object image G that has not been converted to grayscale. In this case, a distribution in a color space (e.g., a three-dimensional space formed by axes corresponding to each of RGB) can be used as the distribution of pixel values ​​(e.g., RGB values) of pixels constituting the object image G. For example, in this color space, the pre-processing execution determination unit 13 can divide the pixels constituting the object image G into two pixel groups corresponding to the ink color and the background color, and determine whether or not the distance in the color space as the difference between the representative pixel values ​​(e.g., RGB values) of the two pixel groups is equal to or greater than a threshold value corresponding to a first threshold value. This makes it possible to determine whether or not to perform pre-processing on the object image G. Furthermore, when preprocessing is performed on the object image G that has not been converted to grayscale, the preprocessing unit 14 can perform the preprocessing by adding or subtracting a uniform value to or from each of the pixels constituting at least one of the two pixel groups. More specifically, the preprocessing unit 14 can perform the preprocessing by, for example, adding or subtracting a uniform value Xr to or from the R value in the RGB values, adding or subtracting a uniform value Xg to or from the G value, and adding or subtracting a uniform value Xb to or from the B value.

[0065] The individual identification system 1 and the information processing device 10 in the above-mentioned embodiment may be realized in whole or in part by a computer. In that case, a program for realizing this function may be recorded in a computer-readable recording medium, and the program recorded in the recording medium may be read into a computer system and executed to realize the function. The term "computer system" as used herein includes hardware such as an OS and peripheral devices. The term "computer-readable recording medium" refers to portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, and storage devices such as hard disks built into a computer system. The term "computer-readable recording medium" may also include a medium that dynamically holds a program for a short period of time, such as a communication line when a program is transmitted via a network such as the Internet or a communication line such as a telephone line, and a medium that holds a program for a certain period of time, such as a volatile memory inside a computer system that is a server or client in that case. The above-mentioned program may be a program for realizing a part of the above-mentioned function, or may be a program that can realize the above-mentioned function in combination with a program already recorded in the computer system, or may be a program that is realized using a programmable logic device such as an FPGA. [Explanation of symbols]

[0066] 1. Individual Identification System 10...Information processing device 11...Image acquisition section 12…Grayscale conversion section 13...Pre-processing execution determination unit 14...Pretreatment section 15…Individual Identification Section 16...Output section 17...Image information storage section 18…Feature point memory section 20...Imaging device 30…Display device

Claims

1. an image acquisition unit that acquires an object image of an object to be individually identified among the printed matter; a pre-processing execution determination unit that determines whether or not to perform pre-processing on the object image based on a distribution of pixel values ​​of each of the pixels that constitute the object image; a pre-processing unit that changes pixel values ​​of the object image by adding or subtracting a uniform offset value to each pixel value of a pixel group of pixels that constitute the object image and whose pixel values ​​are within a specific range when performing pre-processing on the object image; An individual identification system comprising:

2. the pre-processing execution determination unit determines that pre-processing is to be performed on the object image when pixels constituting the object image are divided into two pixel groups according to a frequency of the distribution, and a difference between representative pixel values ​​of the two pixel groups is smaller than a first threshold value; The individual identification system according to claim 1 .

3. when the pixels constituting the object image are divided into three or more pixel groups according to the frequency of the distribution, the pre-processing execution determination unit regards two adjacent pixel groups, each of which has a difference between their representative pixel values ​​that is less than a second threshold value, as one pixel group; The individual identification system according to claim 2 .

4. the preprocessing unit changes pixel values ​​of the object image so that a difference between respective representative pixel values ​​of two pixel groups becomes greater than the first threshold value. The individual identification system according to claim 2 .

5. the pre-processing unit adds or subtracts a uniform offset value to each pixel value of a pixel included in at least one of the two pixel groups, thereby changing the pixel values ​​of the object image so that a difference between the representative pixel values ​​of each of the two pixel groups becomes larger than the first threshold value. The individual identification system according to claim 2 .

6. An individual identification method performed by a computer that performs individual identification, An object image is obtained by capturing an image of an object to be individually identified among the printed matter; determining whether to perform pre-processing on the object image based on a distribution of pixel values ​​of the pixels constituting the object image; When performing pre-processing on the object image, a uniform offset value is added to or subtracted from each pixel value of a pixel group of pixels constituting the object image, the pixel values ​​of which are within a specific range, thereby changing the pixel values ​​of the object image. Individual identification method.

7. A computer that performs an individual identification system acquiring an image of an object to be individually identified among the printed matter; determining whether or not to perform pre-processing on the object image based on a distribution of pixel values ​​of the pixels constituting the object image; When performing pre-processing on the object image, pixel values ​​of the object image are changed by adding or subtracting a uniform offset value to each pixel value of a pixel group that includes pixels constituting the object image and whose pixel values ​​fall within a specific range. program.

Images