JP7377161B2 - Individual identification system - Google Patents

Description

Embodiments of the present invention relate to an individual identification system that can identify an individual based on an individual's unique pattern, so-called individual fingerprint, appearing on the surface of the individual.

In recent years, for example, industrial products are required to have high traceability not only in the distribution process after production but also in the production process. For example, even parts of the same type produced in the same process and in the same lot may have slight differences in performance that pose no problem in normal use. This is because in the production process of a product, if we can take into account these slight differences in the characteristics of individual parts and manufacture the product with the optimal combination, we can improve the performance of that product. .

Furthermore, in conventional production processes, processing conditions for parts, etc. and inspection results are not linked, making it difficult to reflect inspection results in processing conditions, etc. On the other hand, if it is possible to track each part during the production process, inspection results can be appropriately reflected in processing conditions, etc., and as a result, defects can be reduced and quality and performance can be further improved. .

2. Description of the Related Art Conventionally, methods for tracking and managing products and the like include methods that use, for example, two-dimensional codes and RFID. However, two-dimensional codes and RFIDs are expensive to print and paste. Therefore, it is economically difficult to attach a two-dimensional code or RFID to each of the many parts that make up a product.

Therefore, a technique for identifying individuals using the surface shape unique to each individual is attracting attention. If the surface shape or pattern of an object is like this, there is no need to attach a two-dimensional code, RFID, etc., so tracking and management of each part can be realized at low cost.

Japanese Patent Application Publication No. 2012-103758 Japanese Patent Application Publication No. 2012-181566

However, inside a factory, dust, lighting inside the factory, natural light such as outside light, noise sources in the surroundings, etc. become disturbance factors, and there is a possibility that the shooting environment between capturing the registered image and capturing the identification image will change significantly. There is. Then, the extracted feature points, local features, etc. may change, and the identification performance may deteriorate.

The present invention has been made in view of the above-mentioned problems, and its purpose is to set appropriate conditions according to changes in the environment, reduce the influence of external disturbances, and appropriately highlight the desired characteristic pattern. An object of the present invention is to provide an individual identification system that can improve identification accuracy.

The individual identification system (1) of the embodiment is an image acquisition system capable of executing an image acquisition process of acquiring images of an identification target photographed by photographing devices (10a, 10b) on the registration side and identification side as registered images and identification images, respectively. a processing unit (212), a feature point extraction processing unit (221) capable of executing a feature point extraction process for extracting feature points included in the registered image and the identification image acquired in the image acquisition process, and the feature point a local feature calculation processing unit (222) capable of executing a local feature calculation process that calculates a local feature of the feature point extracted in the extraction process; and a photography target that adjusts photography target conditions that are conditions related to the photography target. a photographing target condition adjustment processing unit (312) capable of executing a condition adjustment process; and a photographing equipment condition adjustment processing unit (311) capable of executing a photographing equipment condition adjustment process for adjusting photographing equipment conditions that are conditions related to the photographing equipment. , is provided.

According to this, the user operates the imaging target condition adjustment processing unit (312) and the imaging equipment condition adjustment processing unit (311) to set the imaging target conditions suitable for extraction of feature points and calculation of local feature quantities. Shooting equipment conditions can be adjusted. Therefore, by setting appropriate conditions according to changes in the environment, it is possible to reduce the influence of external disturbances, and it is possible to appropriately highlight the desired characteristic pattern. As a result, identification accuracy can be improved.

A block diagram showing an example of a schematic configuration of an individual identification system according to an embodiment A block diagram showing an example of a schematic configuration of a photographing device included in an individual identification system according to an embodiment A block diagram showing an example of a schematic configuration of a feature extraction device included in an individual identification system according to an embodiment A block diagram showing an example of a schematic configuration of a registration processing device included in an individual identification system according to an embodiment A block diagram showing an example of a schematic configuration of a recording device included in an individual identification system according to an embodiment A block diagram showing an example of a schematic configuration of a dictionary device included in an individual identification system according to an embodiment A block diagram showing an example of a schematic configuration of an image matching identification device included in an individual identification system according to an embodiment A block diagram showing an example of a schematic configuration of a photographing condition adjustment device included in an individual identification system according to an embodiment A block diagram showing an example of a schematic configuration of a verification identification condition adjustment device included in an individual identification system according to an embodiment A block diagram showing an example of a schematic configuration of a visualization device included in an individual identification system according to an embodiment Diagram showing an example of execution results of feature point extraction processing in an individual identification system according to an embodiment (Part 1) Diagram showing an example of the execution result of feature point extraction processing in the individual identification system according to one embodiment (Part 2) A diagram conceptually illustrating an example of a method for calculating global features from local features in an individual identification system according to an embodiment A diagram conceptually illustrating an example of a method for generating a global feature dictionary in an individual identification system according to an embodiment A diagram conceptually illustrating an example of narrowing down processing using global features in an individual identification system according to an embodiment A diagram conceptually illustrating an example of identification processing using local features in an individual identification system according to an embodiment A diagram showing an example of a parameter set in an individual identification system according to an embodiment A diagram conceptually showing an example of the number of positive correspondence points and the number of negative correspondence points in an individual identification system according to an embodiment A diagram showing an example of a combination of conditions generated by round-robin condition generation processing in an individual identification system according to an embodiment A diagram showing an example of a graph generated in optimal condition presentation processing in an individual identification system according to an embodiment A diagram showing an example of display content displayed on a display device in feature point display processing in an individual identification system according to an embodiment Diagram (Part 1) showing an example of display content displayed on a display device in matching result display processing in an individual identification system according to an embodiment Diagram (Part 2) showing an example of display content displayed on a display device in matching result display processing in an individual identification system according to an embodiment

An embodiment will be described below with reference to the drawings. Note that the arrows in each block diagram indicate the flow of data. Each of the processing units or functional blocks described in this embodiment may be configured with different hardware, or a single piece of hardware may share a plurality of devices or processing units. Further, although details are not shown, the hardware constituting each device and each processing unit is mainly composed of a microcomputer having a CPU, ROM, RAM, and a storage area such as a rewritable flash memory. be able to. The storage area stores individual identification programs for realizing each device and each processing section. By executing the individual identification program in the CPU, each process in each device and each processing unit can be realized.

That is, each processing unit described in this embodiment is executed by the CPU executing a computer program stored in a non-transitional physical storage medium such as the above-mentioned storage area and executing processing corresponding to the computer program. It is realized, that is, it is realized by software. Note that a configuration may be adopted in which at least a part of each processing unit is realized by hardware.

The individual identification system 1 of this embodiment compares a plurality of pre-registered images with an identification image taken with a photographing device, and identifies a registered image in which the subject of the identification image is captured from among the plurality of registered images. It is a system that allows you to This individual identification system 1 can be used, for example, to trace parts in a factory or the like, but is not limited thereto.

As shown in FIG. 1, the individual identification system 1 includes a photographing device 10, an input device 11, a display device 12, an individual identification device 20, and an adjustment device 30. In the case of this embodiment, the individual identification system 1 includes a registration-side imaging device 10 and an identification-side imaging device 10. In the following description, when these are to be distinguished, the registration-side imaging device 10 may be referred to as the registration-side imaging device 10a, and the identification-side imaging device 10 may be referred to as the identification-side imaging device 10b. In this embodiment, the registration-side photographing device 10a photographs an image for registration in the individual identification device 20. Further, the identification-side photographing device 10b is for photographing the subject to be identified. In the present embodiment, the registered image mainly means an image that is photographed by the registration-side photographing device 10a and is registered or has been registered in the individual identification device 20. Moreover, the identification image mainly means an image photographed by the photographing device 10b on the identification side.

The photographing device 10 includes a camera 101, a light 102, and a stage 103. The camera 101 has a function of photographing a subject such as a workpiece. The camera 101 is configured to be able to change shutter speed, image sensor sensitivity, resolution, photographic range, focus, etc. in accordance with instructions from the outside. The illumination 102 is for illuminating the subject, and is configured to be able to change brightness, luminous intensity, color, etc. according to instructions from the outside. The stage 103 supports the camera 101 and the lighting 102, and also has a subject placed thereon, and is configured to be able to change the distance and angle, that is, the posture, between the camera 101, the lighting 102, and the subject.

The input device 11 is an interface that accepts user operations, and is, for example, a keyboard, a mouse, or the like. The user can perform various settings for the adjustment device 30 and the individual identification device 20 by operating the input device 11. The display device 12 is a display capable of displaying characters, images, etc., and can display information received from the individual identification device 20 and the adjustment device 30. By viewing the information displayed on the display device 12, the user can confirm the results of processing by the individual identification device 20 and the adjustment device 30, the settings, etc.

The individual identification device 20 includes a photographing device 21, a feature extraction device 22, a registration processing device 23, a recording device 24, a dictionary device 25, an image matching and identification device 26, a photographing condition providing device 27, and a feature extraction condition providing device 28. ing. In the case of this embodiment, the imaging device 21, feature extraction device 22, registration processing device 23, recording device 24, imaging condition provision device 27, and feature extraction condition provision device 28 are provided on the registration side and the identification side, respectively. There is. In this case, the device surrounded by the broken line a becomes the device on the registration side, and the device surrounded by the broken line b becomes the device on the identification side. Note that the registration side device a and the identification side device b may be configured with different hardware, or may be configured with the same hardware. Further, the individual identification device 20 and the adjustment device 30 may be installed on a cloud server connected to a telecommunications line such as the Internet.

The photographing device 21 has a function of controlling the driving of the photographing device 10 and photographing and acquiring an image of a subject. As shown in FIG. 2, the photographing device 21 includes a photographing processing section 211 and an image acquisition processing section 212. The photographing processing unit 211 drives the photographing device 10 under predetermined conditions to photograph a subject. In this case, the predetermined conditions when photographing with the photographing device 10 are referred to as photographing conditions. The photographing conditions include, for example, photographing device conditions that are conditions regarding the photographing device 10, and photographing target conditions that are conditions regarding the subject to be photographed.

The photographing equipment conditions are mainly conditions related to optical equipment, such as the shutter speed of the camera 101 included in the photographing equipment 10, the sensitivity, resolution, photographing range, focus, magnification, aperture of the image sensor, the type and brightness of the illumination 102, etc. There are luminosity, color, etc. The photographing target conditions are conditions related to the photographing target, that is, the subject, and include, for example, the posture of the subject, that is, the angle of the subject with respect to the camera 101 or the illumination 102, the distance between the camera 101 or the illumination 102, and the subject, and the like.

The photographing processing unit 211 can change the photographing device conditions and photographing target conditions based on the instruction content received from the photographing condition providing device 27 and drive the photographing device 10 to photograph the subject.

The image acquisition processing unit 212 can execute image acquisition processing. The image acquisition process includes a process of acquiring an image of a subject photographed by the photographing device 10 and passing it to the feature extraction device 22. In the case of this embodiment, the image acquisition processing unit 212 on the registration side a passes an image photographed by the imaging device 10a to the feature extraction device 22 on the registration side a as a registered image. Further, the image acquisition processing unit 212 on the identification side b passes the image photographed by the imaging device 10b to the feature extraction device 22 on the identification side b as an identification image. In this embodiment, among the image acquisition processes, the process of acquiring a registered image performed on the registration side a is referred to as a registered image acquisition process, and the process of acquiring an identification image performed on an identification side b is referred to as an identification image acquisition process. Sometimes. Note that the image acquisition process may include a process of acquiring an image from a means other than the imaging device 10, such as acquiring an image via the Internet or a storage medium.

The feature extraction device 22 has a function of extracting features included in the image acquired from the imaging device 21. The feature extraction device 22 has a function of calculating feature points, local features, and global features as image features. As shown in FIG. 3, the feature extraction device 22 includes a feature point extraction processing section 221, a local feature calculation processing section 222, a local feature group classification processing section 223, and a global feature calculation processing section 224. . The feature extraction device 22 receives feature extraction conditions from the feature extraction condition applying device 28 and executes processing in each of the processing units 221 to 224. The feature extraction conditions include, for example, the number of steps of the pyramid image when extracting feature points and calculating local feature amounts, and thresholds for determining the likelihood of corners or curves.

The feature point extraction processing unit 221 can execute feature point extraction processing. The feature point extraction process includes a process of extracting one or more feature points P appearing on the surface of the subject S from the image acquired by the image acquisition processing unit 212, as shown in FIG. In the case of this embodiment, the feature point extraction process extracts, for example, several thousand feature points from one image. In this case, the feature point extraction processing unit 221 on the registration side a can extract the feature points of the registered image photographed by the imaging device 10a on the registration side. Further, the feature point extraction processing unit 221 on the identification side b can extract feature points of the identification image captured by the imaging device 10b on the identification side.

Here, as shown in FIG. 12, for example, by extracting non-geometric patterns such as corners and curves among the patterns appearing on the surface of the workpiece S as feature points P, It is thought that it is possible to identify individual parts even among parts of the same type. This is because, for example, geometric patterns such as straight lines and right angles that appear on the surface of parts are often patterns that appear in common across all parts of that type, and are characteristics that can be used to identify individuals among parts. It is difficult to become. On the other hand, non-geometric patterns such as corners and curves are often the result of very slight differences in conditions during the manufacturing process, and are thought to be unique to each individual product. Therefore, the feature point extraction processing unit 221 of this embodiment extracts, as feature points P, patterns of corners and curves that appear locally on the surface of the subject.

The feature point extraction processing unit 221 of this embodiment uses the brightness gradient intensity distribution obtained from the image of the subject to extract corners and curves that appear locally on the surface of the subject as feature points. An image has a brightness gradient and its intensity between pixels. In the case of a uniform pattern, the spread of the brightness gradient strength is small, and in the case of a linear pattern, the brightness gradient strength spreads in a specific direction. On the other hand, in the case of a pattern with corners or curves, the brightness gradient strength spreads in multiple directions. Therefore, the feature point extraction processing unit 221 can extract corners and curves that serve as feature points by searching for a pattern in which the intensity of the brightness gradient spreads in multiple directions.

The local feature amount calculation processing unit 222 is capable of executing local feature amount calculation processing . The local feature amount calculation process includes a process of calculating the feature amount of the feature point as a local feature amount based on the brightness gradient of the surrounding area including the feature point extracted in the feature point extraction process. The local feature has the coordinates of a feature point and a 256-bit binary feature. The existing technology SIFT can be used to calculate the feature amount. Furthermore, if possible, other image feature calculation algorithms such as SURF may be used. In the feature point extraction process, a pyramid image is obtained by reducing the image in multiple stages, and feature points may be extracted independently from each scaled image under the same conditions. Then, the local feature amount calculation processing unit may calculate the local feature amount of the feature points obtained in each pyramid image. By doing so, it is possible to obtain local features that have scale invariance and are not easily affected by differences in scale.

The local feature group classification processing unit 223 can perform local feature group classification processing. The local feature group classification process is a preparatory process for calculating global features of registered images and identification images. In the local feature group classification process, for example, as shown in FIGS. 13A and 13B, a plurality of local features A obtained by performing local feature calculation processing on each image are classified into local feature values. It includes a process of classifying into a predetermined number of local feature groups B, for example, 64 to 4096, according to the value of A. In other words, the local feature amount group classification process includes a process of clustering a large number of local feature amounts A obtained from each image by gathering those whose values are similar. Each element Bn constituting the local feature group B has at least one pair of a feature point obtained from an image and its local feature. Note that n in Bn is an integer indicating the number of elements.

The local feature amount group classification process includes a process of classifying the local feature amount A obtained from each image based on the dictionary information d. The dictionary information d is representative of the feature amounts of the feature points obtained from the learning images acquired in advance, and is prepared before execution of the local feature classification process. The local feature group classification processing unit 223 classifies the local feature A to be classified by applying it to dictionary information d having a feature similar to the local feature A.

The global feature amount calculation processing unit 224 is capable of executing global feature amount calculation processing. The global feature calculation process includes a process of calculating global features based on each local feature group B obtained in the local feature separation process. In the global feature calculation process, as shown in FIG. 13(C), the frequency distribution, that is, the histogram, obtained by voting on the dictionary information d that has a high correlation with each local feature A obtained from the image is calculated as the global feature. Including processing. That is, for the global feature amount, the horizontal axis represents the local feature group B, and the vertical axis represents the number of votes cast for each local feature group B. By using the global feature amount calculated in this way, one image can be represented by one feature amount. For the global feature amount calculation process, for example, VLAD or BAG-OF-WORDS can be adopted as a calculation method.

なお、（１）式のＮＮ（ｖ）＝ｄｋは、画像中に含まれる特徴量の中で、この場合６４個ある各辞書情報ｄと最も相関が高い局所特徴量ｖを意味する。 When VLAD is adopted as the global feature calculation method, the global feature calculation process is performed by, for example, adding a vector V, which is the sum of the difference vectors between the local feature and a representative value, which is dictionary information that has a high correlation, for the number of representative values. A vector that connects them is calculated as a global feature vector V. Specifically, the global feature amount calculation processing unit 224 calculates the vector V of the global feature amount by applying the local feature amount obtained from the image to the calculation of the following equation (1). In this case, the global feature vector V is composed of 64 128-dimensional vectors. In this case, let v be the local feature included in each local feature group B, and let d be the dictionary information. Furthermore, if the number of features included in each local feature group B is N, then 1≦i≦N.

Note that NN(v)=dk in equation (1) means the local feature amount v that has the highest correlation with each of the 64 pieces of dictionary information d among the feature amounts included in the image.

Furthermore, when BAG-OF-WORDS is adopted as the global feature, the global feature dictionary generation process calculates the global feature as follows. In this case, the global feature amount is a vector with dimensions equal to the number of dictionary information, for example, a 64-dimensional vector V. The global feature dictionary generation processing unit 251 searches for the nearest dictionary information d for each feature using dictionary information serving as 64 representative values, and adds 1 to the element of the vector V that is the closest. As a result, the vector V becomes a histogram of feature quantities, and the sum of the elements of the vector V has the same number of feature quantities. This vector V is used by dividing the whole by this total number and normalizing it.

The registration processing device 23 has a function of recording data obtained in each process in the recording device 24. As shown in FIG. 4, the registration processing device 23 includes a search target image registration processing section 231, an evaluation image registration processing section 232, a local feature registration processing section 233, and a global feature registration processing section 234. . The search target image registration processing unit 231 is capable of executing search target image registration processing. The search target image registration process includes a process of registering in advance a plurality of search target images as registered images in the search target image registration unit 241 of the recording device 24.

The evaluation image registration processing unit 232 is capable of executing evaluation image registration processing. The evaluation image registration process includes a process of registering an image used when evaluating the matching process executed by the image matching and identification device 26 as an evaluation image in the evaluation image registration section 242 of the recording device 24. The evaluation image includes a registered image group including a plurality of registered images, each of which is taken of a different subject, and an identification image of the same subject as one of the registered image groups, but taken under different conditions. has been done. In this case, there may be a plurality of identification images depending on the number of subjects and conditions. In addition, the registered image and the identification image taken of the same subject are registered as images for evaluation in the recording device 24 in a state in which they are linked to each other by, for example, being assigned the same or partially common identification number. It is registered in the section 242.

The local feature registration processing unit 233 can execute local feature registration processing. The local feature amount registration process includes a process of registering the local feature amount of the feature point extracted from the registered image by the process performed by the feature extraction device 22 in the local feature amount registration unit 243 of the recording device 24. Further, the local feature amount registration process includes a process of registering the local feature amount of the feature point extracted from the evaluation image by the process performed by the feature extraction device 22 in the local feature amount registration unit 243 of the recording device 24. The local feature registration processing unit 233 registers the local feature obtained from each image in association with the extraction source image.

The global feature registration processing unit 234 is capable of executing global feature registration processing. The global feature amount registration process includes processing to register the global feature amount related to the registered image in the global feature amount registration unit 244 of the recording device 24. Further, the global feature amount registration process includes processing for registering the global feature amount related to the evaluation image in the global feature amount registration unit 244 of the recording device 24. The local feature registration processing unit 233 registers the global feature obtained from each image in association with the image from which it is extracted.

The recording device 24 has a function of recording various data sent through each registration process of the registration processing device 23, and constitutes a database. The recording device 24 can also be configured, for example, by a recording device of a computer that constitutes the individual identification device 20, an external server connected via a network line, or the like. As shown in FIG. 5, the recording device 24 includes a search target image registration section 241, an evaluation image registration section 242, a local feature registration section 243, a global feature registration section 244, and a global feature dictionary registration section 245. are doing.

The search target image registration unit 241 is a recording area in which the search target image sent from the search target image registration processing unit 231 of the registration processing device 23 is recorded. The evaluation image registration section 242 is a recording area in which the evaluation image sent from the evaluation image registration processing section 232 of the registration processing device 23 is recorded. The local feature amount registration unit 243 is a recording area that records the local feature amount sent from the local feature amount registration processing unit 233 of the registration processing device 23. The global feature amount registration unit 244 is a recording area that records the global feature amount sent from the global feature amount registration processing unit 234 of the registration processing device 23.

Further, the global feature dictionary registration unit 245 is a recording area that records the global feature dictionary sent from the global feature dictionary registration processing unit 252 of the dictionary device 25. Before the image matching and identification device 26 executes the search process, a plurality of global feature dictionaries having elements of a plurality of dictionary information calculated based on local features obtained from pre-obtained learning images are registered. ing.

The dictionary device 25 has a function of generating a plurality of pieces of dictionary information based on feature points extracted from a plurality of images and their local features, and generating, registering, and setting a global feature dictionary based on the dictionary information. has. As shown in FIG. 6, the dictionary device 25 includes a global feature dictionary generation processing section 251 and a global feature dictionary registration processing section 252.

The global feature dictionary generation processing unit 251 is capable of executing global feature dictionary generation processing. The global feature dictionary generation process includes, for example, a process of generating a global feature dictionary whose dictionary information includes representative values of each of the local feature groups obtained from a plurality of learning images. In this case, the global feature dictionary generation process includes a process of setting the centroid of the local feature group as a representative value in the dictionary information. Dictionary information is an element constituting a global feature dictionary, and a plurality of pieces of dictionary information are collected to form one global feature dictionary.

In this case, for example, in the global feature dictionary generation process, the global feature dictionary generation processing unit 251 can generate the global feature dictionary as follows. That is, in the global feature dictionary generation process, the global feature dictionary generation processing unit 251 first generates m learning images calculated by the feature extraction device 22, as shown in FIGS. 14(A) and 14(B). Receive local feature amounts A from T1 to Tm. Next, as shown in FIG. 14C, the global feature dictionary generation processing unit 251 uses, for example, the k-means method to convert the received local feature A into k clusters, in this case 64 clusters, that is, local feature amounts. It is classified into groups B1 to Bk, and centroids C1 to Ck are determined for each of the local feature groups B1 to Bk. Then, the global feature dictionary generation processing unit 251 uses the centroids C1 to Ck of each of the local feature groups B1 to Bk obtained in this way to the dictionary information d1 as representative values of each of the local feature groups B1 to Bk. ~dk, and generates a global feature dictionary E that compiles the dictionary information d1~dk. Then, the global feature dictionary generation processing unit 251 further generates a plurality of global feature dictionaries E using different learning images T. Note that the dictionary information d1 to dk are not limited to centroids as long as they are values representative of each of the local feature groups B1 to Bk.

The global feature dictionary registration processing unit 252 is capable of executing global feature dictionary registration processing. The global feature dictionary registration process includes a process of recording a plurality of global feature dictionaries Eic generated in the global feature dictionary generation process in the global feature dictionary registration unit 245 of the recording device 24. In this case, the global feature dictionary E includes a plurality of pieces of dictionary information d1 to dk that constitute the global feature dictionary E.

The image comparison and identification device 26 has a prominent function of comparing the registered image recorded in the recording device 24 with the identification image output from the feature extraction device 22 on the identification side and identifying whether the images match or do not match. The image matching and identification device 26 includes a search processing section 261, a matching processing section 262, and a result notification processing section 263, as shown in FIG.

The search processing unit 261 is capable of executing search processing. The search process is performed using the global feature amount related to the identification image sent from the feature extraction device 22 on the identification side, among the global feature amounts related to the plurality of registered images recorded in the search target image registration unit 241 of the recording device 24. Includes processing to search for items with high correlation. In the case of this embodiment, the search processing section 261 includes a narrowing down processing section 2611 that can perform narrowing down processing, and a specific processing section 2612 that can perform specific processing. That is, in the case of this embodiment, the search process includes a narrowing process and a specifying process.

As shown in FIG. 15, the narrowing down process executed by the narrowing down processing unit 2611 selects the global feature amount and the amount of features that the identified image has from among the plurality of registered images registered in the search target image registration unit 241 of the recording device 24. The process includes a process of narrowing down candidates to a predetermined number of registered images having global feature quantities with high correlation. If the number of registered images registered in the search target image registration unit 241 is, for example, several thousand to several hundred thousand, the registered image candidates are narrowed down to, for example, several to several tens of images by the narrowing down process.

The identification processing performed by the identification processing unit 2612 can be performed using an image matching method using local feature amounts and an algorithm that verifies the consistency of the geometric arrangement of feature points. The identification processing unit 2612 first compares the local feature amount of the candidate registered image extracted by the narrowing down process with the local feature amount of the identification image, and pairs feature points with the minimum difference in feature amount from both images. Find it as. Then, the identification processing unit 2612 extracts feature points whose relative positional relationships with other feature points do not contradict each other as corresponding points, and identifies the registered image with the largest number of corresponding points.

In this case, as shown in FIG. 15, the narrowing down process is a process that compares the respective global features of the registered image and the identified image, so it is better than a specific process that compares all the local features included in the image. Although the accuracy is low, the amount of processing is small and processing can be performed at high speed. On the other hand, as shown in FIG. 16, the identification process is a process of searching for corresponding points for all of the local feature amounts A of each of the registered image and the identification image, so it is a narrowing down process that compares the global feature amounts with each other. Although it requires more processing and takes more time, it has better accuracy. In this manner, the search process performs two steps of narrowing down and specifying, thereby making it possible to quickly and accurately specify a registered image that is the same as the identification image from among a plurality of registered images.

The verification processing unit 262 is capable of performing verification processing. The matching process is a process that does not involve searching for images, and is executed when the adjustment device 30 matches specific images to adjust various conditions, algorithms, parameters, and the like. The matching process is performed using the same algorithm as the algorithm for matching images in the search process. Therefore, the matching processing section 262 can be a part of the search processing section 261. Various conditions, algorithms, parameters, etc. adjusted by the adjustment device 30 are reflected in the search processing by the search processing unit 261.

The matching process includes, for example, comparing the local feature amounts of a specific registered image registered in the evaluation image registration unit 242 as an evaluation image with the identification image, obtaining the number of corresponding feature points, and identifying the registered image and the identified image. Includes processing to match the image. In this case, the matching processing unit 262 can match the identification image with a registered image taken of the same subject based on an instruction from the adjustment device 30. Further, the matching processing unit 262 can also match registered images taken of different subjects and identification images based on instructions from the adjustment device 30. The matching processing unit 262 outputs the number of corresponding points of local feature amounts between the registered image and the identification image.

Further, the matching processing unit 262 can execute matching processing using the parameter set received from the matching identification condition adjustment device 32. For example, as shown in FIG. 17, the parameter set is a combination of values that can be set for a plurality of parameters for setting conditions for search processing and matching processing. Condition 1, Condition 2, Condition 3, etc. in FIG. 17 each constitute one parameter set. In the example of FIG. 17, there are a total of 5400 parameter sets.

The result notification processing unit 263 can execute result notification processing. The result notification process includes a process of outputting the results of the search process and the results of the matching process to the adjustment device 30 or an external device. In this case, the search processing result includes information regarding the specified registered image. Furthermore, the result of the matching process includes the number of corresponding points.

The photographing condition providing device 27 has a function of receiving instructions from the photographing condition adjusting device 31 of the adjusting device 30 and applying the above-described photographing conditions to the photographing device 21 . The feature extraction condition provision device 28 has a function of receiving an instruction from the matching identification condition adjustment device 32 of the adjustment device 30 and provides a feature extraction condition to the feature extraction device 22 .

The adjustment device 30 has a function of visualizing the results obtained from the image matching and identification device 26 and displaying it on the display device 12, and receiving feedback from the image matching and identification device 26, and adjusts the imaging conditions of the imaging device 10 and the feature extraction device 22. It has a function to adjust various conditions, algorithms, parameters, etc. used in the individual identification device 20, such as conditions and algorithms used in the individual identification device 20. As shown in FIG. 1, the adjustment device 30 includes a photographing condition adjustment device 31, a verification and identification condition adjustment device 32, and a visualization device 33.

The photographing condition adjustment device 31 automatically adjusts various conditions for photographing the photographing target using the photographing device 10 automatically in response to feedback from the image matching and identification device 26 or manually in response to operation from the user. It has a function of instructing the imaging condition providing device 27 about the conditions. For example, by operating the input device 11, the user can adjust various conditions through the processing of the imaging condition adjustment device 31. Further, the imaging condition adjustment device 31 has a function of automatically adjusting various conditions according to the number of feature points obtained by the processing of the feature extraction device 22. As shown in FIG. 7, the photographing condition adjustment device 31 includes a photographing equipment condition adjustment processing section 311, a photographing target condition adjustment processing section 312, a brute force condition generation processing section 313, and an optimal condition presentation processing section 314. .

The imaging equipment condition adjustment processing unit 311 is capable of executing imaging equipment condition adjustment processing for adjusting imaging equipment conditions. By executing the photographing equipment condition adjustment process, the conditions mainly related to the optical equipment, such as the shutter speed of the camera 101 included in the photographing equipment 10, the sensitivity, resolution, photographing range, focus of the image sensor, the brightness and luminous intensity of the illumination 102, color, etc. is adjusted. The photographing equipment condition adjustment processing unit 311 transmits the adjusted photographing equipment conditions to the photographing condition imparting device 27. Thereby, the photographing device 21 operates mainly the camera 101 and the lighting 102 of the photographing device 10 in accordance with the photographing device conditions adjusted by the photographing device condition adjustment processing section 311 to photograph the object.

The photographing target condition adjustment processing unit 312 is capable of executing photographing target condition adjustment processing that adjusts photographing target conditions, which are conditions related to a subject to be photographed. By executing the photographing target condition adjustment process, for example, the posture of the subject, that is, the angle of the subject with respect to the camera 101 or the illumination 102, the distance between the camera 101 or the illumination 102, and the subject, etc. are adjusted. The photographing target condition adjustment processing section 312 transmits the adjusted photographing target conditions to the photographing condition imparting device 27. Thereby, the photographing device 21 operates mainly the stage 103 of the photographing device 10 in accordance with the photographing target conditions adjusted by the photographing target condition adjustment processing unit 312 to photograph the photographing target.

The brute force condition generation processing unit 313 is capable of executing brute force condition generation processing. The brute force condition generation process includes a process of brute force generating a combination of photographing target conditions within a preset range and photographing equipment conditions within a preset range. In this case, the image acquisition process executed by the image acquisition processing unit 212 of the photographing device 21 includes the process of acquiring images photographed under all the conditions generated in the round-robin condition generation process. Thereby, the image acquisition processing unit 212 can acquire a large number of images shot under different conditions, that is, a large number of images shot under different shooting target conditions and shooting equipment conditions.

For example, in the example of FIG. 19, the illumination distance, which is the distance between the subject and the illumination 102, is set as the photographing target condition, and the light amount setting value, which is the light amount of the illumination 102, is set as the photographing equipment condition. In this case, the brute force condition generation processing unit 313 generates all possible combinations of the illumination distance and light amount setting value by brute force. Then, the image acquisition processing unit 212 automatically captures images under the conditions of this round-robin combination, and acquires a registered image and an identification image.

The optimal condition presentation processing unit 314 can execute optimal condition presentation processing. In the optimal condition presentation process, the matching processing unit 262 performs a matching process on the registered images and identification images taken of the same subject among the registered images and identification images taken under all the conditions generated in the round-robin condition generation process. The combination of the photographing target condition and the photographing equipment condition that has a large number of corresponding points obtained through the matching process is presented to the user as the optimum condition, or is automatically set. In this case, the optimal condition presentation processing section 314 may display the optimal conditions on the display device 12. This allows the user to easily find and set the optimal combination of the imaging equipment conditions and the imaging target conditions.

For example, when the registered image and the identification image are images of the same subject taken under the conditions shown in FIG. 19, the optimal condition presentation processing unit 314 calculates the graph shown in FIG. In this case, a graph is generated in which the horizontal axis represents the light intensity setting value, ie, Lux, and the vertical axis represents the number of corresponding points. Then, the optimal condition presentation processing unit 314 finds a combination of conditions that has a large number of correspondence points between the registered image and the identified image and has high robustness from this graph. In this case, it can be seen that the light intensity setting value at the apex of the graph of A1 with illumination distance of 1 has a large number of corresponding points and has the highest robustness. Therefore, in this case, the optimal condition presentation processing unit 314 presents the light amount setting value at the apex of the graph of A1 at illumination distance 1 and illumination distance 1 as the optimal condition for the combination of the photographing target condition and the photographing equipment condition. In this way, the optimal condition presentation processing unit 314 presents the user with the optimal combination of the imaging target conditions and the imaging equipment conditions. Note that the optimal condition presentation processing unit 314 can also select some candidate conditions with a large number of corresponding points and high robustness, and present these conditions to the user as recommended conditions. Further, the optimal condition presentation processing unit 314 may perform a process of displaying the graph shown in FIG. 20 on the display device 12.

In the optimal condition presentation process, multiple feature extraction conditions, which are conditions related to the feature point extraction process, are prepared in advance, and registered images and identification images taken that are the same under all conditions generated in the round-robin condition generation process are prepared in advance. It includes a process of presenting a combination of conditions that will result in the largest number of corresponding points obtained by executing a matching process using feature points obtained by executing a feature point extraction process under a plurality of feature extraction conditions. Thereby, the user can easily find and set the optimal combination of the imaging equipment conditions and the imaging target conditions, including the feature extraction conditions.

The matching identification condition adjustment device 32 has a function of tuning the parameters of the matching identification algorithm used in the search processing section 261 and the matching processing section 262 of the image matching and identification device 26. The matching and identification condition adjustment device 32 adjusts the conditions and parameters used in the processing of the feature extraction device 22 and the image matching and identification device 26 automatically in response to feedback from the image matching and identification device 26 or manually in response to operation from the user. etc., and instructs the feature extraction device 22 and the image matching and identification device 26 about the conditions and parameters. In the case of this embodiment, the matching identification condition adjustment device 32 performs matching in the matching processing section 262 of the image matching and identification device 26 using the evaluation image registered in the evaluation image registration section 242 of the recording device 24, In this case, the conditions and parameters are adjusted according to the matching result.

The user can adjust the conditions and parameters used by the feature extraction device 22 and the image matching and identification device 26 by operating the input device 11 and executing processing in the matching and identification condition adjustment device 32, for example. As shown in FIG. 9, the matching identification condition adjustment device 32 includes a feature extraction condition adjustment processing section 321, a global feature dictionary setting processing section 322, a global feature correlation coefficient calculation processing section 323, a parameter set generation processing section 324, and a parameter set setting processing section 325.

The feature extraction condition adjustment processing unit 321 can execute feature extraction condition adjustment processing. The feature extraction condition adjustment processing includes processing for adjusting feature extraction conditions used in the feature point extraction processing in the feature point extraction processing section 221 or the local feature amount calculation processing in the local feature amount calculation processing section 222. The user operates the input device 11 to adjust and set feature extraction conditions such as the number of stages of the pyramid image when extracting feature points and calculating local feature amounts, and thresholds for determining the likeness of corners and curves. be able to.

The feature extraction condition adjustment process is based on the result of the matching process in the matching processing unit 262 of the image matching and identification device 26, so that the ratio between the number of positive correspondence points and the number of negative correspondence points is maximized for the feature extraction condition by the local feature amount calculation process. It includes processing to adjust to. In this embodiment, as shown in FIG. 18(A), the correct corresponding point refers to the same subject as the subject of the identification image among the registered images registered in the evaluation image registration unit 242 of the recording device 24. It means the corresponding point between the photographed registered image and the identification image. The number of correct corresponding points means the number of correct corresponding points. In the example of FIG. 18A, the subject S1 shown in the registration image and the subject S1 shown in the identification image are the same.

Further, as shown in FIG. 18(B), a negative corresponding point is a registered image obtained by photographing a subject different from the subject of the identification image among the registered image group registered in the evaluation image registration unit 242 of the recording device 24. means the corresponding point with the identification image. The number of negative corresponding points means the number of negative corresponding points. In the example of FIG. 18(B), the subject S2 shown in the registration image and the subject S1 shown in the identification image are different.

The global feature dictionary setting processing unit 322 can execute global feature dictionary setting processing. In the global feature dictionary setting process, a score is calculated for each piece of global feature calculation statistical information, and the global feature dictionary that has the highest score is used in the local feature group classification process of the local feature group classification processing unit 223. This includes the process of automatically setting the . Global feature calculation statistical information is statistical information obtained in the process of calculating global features by global feature calculation processing, that is, in the process of local feature group classification processing in the local feature group classification processing unit 223. It can be expressed by a histogram obtained by voting on dictionary information that has a high correlation with the feature amount or by a score of statistics of correlation values. Dictionary information with a high score based on statistical information such as the number of times specific dictionary information is selected is considered to have high performance. According to the global feature dictionary setting process, a global feature dictionary having such dictionary information with high performance can be automatically selected.

The global feature dictionary setting process selects registered images and identification images of the same subject based on global feature correlation coefficients among a plurality of global feature dictionaries registered in the global feature dictionary registration unit 245. is the same, and selects a global feature dictionary that can determine that the registered image taken of a different subject and the identified image are not the same, and sets it as the global feature dictionary used for local feature group classification processing. including.

The global feature correlation coefficient serves as one index that indicates the degree of matching between the global feature amounts between the registered image and the identified image, that is, the performance. The greater the correlation between the global feature amounts between two images, the higher the possibility that the same subject is included in both images. Therefore, a high correlation of global features between images taken of the same subject means that the degree of agreement between the global features calculated using the global feature dictionary for images taken of the same subject is high. means. In other words, this means that the performance of the global feature dictionary is high among images taken of the same subject.

On the other hand, the smaller the correlation between the global feature amounts between two images, the higher the possibility that the subjects included in the two images are different. Therefore, a low correlation of global features between images of different subjects means that the degree of agreement between the global features calculated using the global feature dictionary for images of different subjects is low. do. In other words, this means that the performance of the global feature dictionary is high between images taken of different subjects.

In the global feature dictionary setting process, for example, a global feature correlation coefficient is calculated as follows, and the performance of the global feature dictionary can be determined based on the global feature correlation coefficient. In other words, in the global feature dictionary setting process, when two global features are vectors V, the absolute value of each is normalized to 1 and the inner product is taken, and the value of the inner product is close to 1. If the value of the inner product is close to -1, it is determined that the identity is low.

In this case, for example, the global feature vector Va of one of the two global features is defined as the following equation (2), the normalized global feature vector Vb is defined as the following equation (3), and d is defined as a natural number. do. In this case, the normalized global feature vector Vb is expressed by the following equation (4) using the global feature vector Va. If the global feature vector Va is expressed by the following equation (5), then the absolute value of the global feature vector Va is expressed by the following equation (6).

If the other global feature vector Wa of the two global features is given by the following equation (7), then the value of the inner product is given by the following equation (8). Then, the global feature dictionary registration unit 245 determines that if the inner product value obtained by equation (8), that is, the global feature correlation coefficient is close to 1, there is a high degree of identity, that is, there is a high possibility that the images are the same subject. If the value of the inner product is close to -1, it is determined that the identity is low, that is, the possibility that they are the same subject is low.

Further, in the global feature dictionary setting process, for example, a global feature correlation coefficient can be calculated as follows, and the performance of the global feature dictionary can be determined based on the global feature correlation coefficient. That is, the global feature correlation coefficient is a value of the distance between vectors of two global features. Then, in the global feature dictionary setting process, if the global feature correlation coefficient, which is the value of the distance between two global feature vectors, is close to 0, it is determined that the identity is high, and if it is large, the identity is low. I judge that.

In this case, for example, when one of the two global feature vectors Va is set to the above equations (2) and (5), and the other global feature vector Wa is set to the above (7) and the following equation (9), The distance between the global feature vectors Va and Wa is expressed by the following equation (10).

Then, the global feature dictionary registration unit 245 determines that if the distance between the vectors Va and Wa obtained by equation (12) is close, that is, close to 0, the identity is high, that is, there is a high possibility that they are the same subject. If the distance between the vectors Va and Wa is large, it is determined that the identity is low, that is, the possibility that they are the same subject is low. Note that at this time, the respective vectors Va and Wa do not necessarily need to be normalized.

The global feature correlation coefficient calculation processing unit 323 is capable of executing global feature correlation coefficient calculation processing. The global feature correlation coefficient calculation process includes processing that compares the global feature of each registered image with the global feature of the identified image, and calculates a global feature correlation coefficient that indicates the correlation between them. For example, in the global feature correlation coefficient calculation process, the global feature correlation coefficient can be calculated by comparing the histograms of the global feature of each registered image and the global feature of the identified image.

The parameter set generation processing unit 324 can execute parameter set generation processing. For example, as shown in FIG. 17, the parameter set generation process includes setting parameters for setting the conditions for the matching process executed by the matching processing unit 262 of the image matching and identification device 26 within the setting range defined for each parameter. It includes a process of generating a plurality of parameter sets by combining each parameter set as a setting value. In the case of this embodiment, the parameter set generation process generates parameter sets for all combinations of setting values of each parameter.

The parameter set setting processing unit 325 can execute parameter set setting processing. The parameter set setting process includes a process of setting a parameter set to be used in the search process by the search processing unit 261 based on the user's operation. The user can, for example, operate the input device 11 to set a parameter set to be used in the search processing by the search processing unit 261.

The visualization device 33 has a function of visualizing and displaying various information on the display device 12, such as information received from the image matching and identification device 26 and information regarding the adjustment results of the imaging condition adjustment device 31 and the matching and identification condition adjustment device 32. have In this embodiment, visualization includes displaying data on the display device 12 not only as characters and numbers but also as figures, graphs, and the like. As shown in FIG. 10, the visualization device 33 includes a feature point display processing section 331, a global feature calculation statistical information display processing section 332, a matching result display processing section 333, and a condition display processing section 334.

The feature point display processing unit 331 can execute feature point display processing. The feature point display process can be executed when various conditions and parameters are adjusted by the photographing condition adjustment device 31 and the verification and identification condition adjustment device 32. In the feature point display process, for example, as shown in FIG. 12. That is, the feature point P is displayed superimposed on the subject S in the image. This allows the user to visually recognize where on the subject S in the image has been extracted as a feature point. In addition, as shown in FIG. 21A, the feature point display processing differs for each local feature group to which the local feature belongs, in this case, for each dictionary information d to which each feature point P is clustered and applied. It includes a process of displaying each feature point P on the display device 12 in a different manner, for example, in a different color, size, or shape. In FIG. 21, the difference in color of each feature point P is expressed by the difference in shading.

The global feature calculation statistical information display processing unit 332 can execute global feature calculation statistical information display processing. The global feature value calculation statistical information display process can be executed when adjusting various conditions and parameters by the imaging condition adjustment device 31 and the matching identification condition adjustment device 32. The global feature calculation statistical information display process includes processing to display global feature calculation statistical information on the display device 12. The global feature amount calculation statistical information is statistical information obtained in the process of calculating a global feature amount by the global feature amount calculation process executed by the global feature amount calculation processing unit 224 of the feature extraction device 22. As shown in FIG. 21(B), the global feature calculation statistical information display process uses the histogram or correlation value statistics scores obtained by voting on the dictionary information d that has a high correlation with each local feature A. This includes a process of displaying the feature amount calculation statistical information G on the display device 12. That is, the global feature amount calculation statistical information G can be a histogram of global feature amounts or a score of statistics of correlation values obtained in the execution process of global feature amount calculation processing.

The matching result display processing unit 333 can execute matching result display processing. The matching result display process can be executed when various conditions and parameters are adjusted by the imaging condition adjusting device 31 and the matching identification condition adjusting device 32. The matching result display process includes processing for collectively visualizing the results of a plurality of matching processes executed by the matching processing section 262 of the image matching and identification device 26 and displaying the results on the display device 12.

The matching result display process includes a process of displaying the matching results obtained by executing the matching process using each parameter set shown in FIG. 17 in a graph as shown in FIG. 22, for example. In this case, the matching result display processing unit 333 performs matching processing not only between images of the same subject but also between images of different subjects. Then, the matching result display processing unit 333 calculates the number of positive correspondence points, which is the number of correspondences between images taken of the same subject, and the number of negative correspondences, which is the number of correspondences between images taken of different subjects, and graphs them. and display it on the display device 12. In the graph of FIG. 22, the horizontal axis represents the condition number, and the vertical axis represents the number of corresponding points. Further, in FIG. 22, F1 shows a graph of the number of positive corresponding points, and F2 shows a graph of the number of negative corresponding points. In the example of FIG. 22, condition No. It can be seen that N has the largest number of positive corresponding points and the smallest number of negative corresponding points. Thereby, the user can grasp at a glance the condition number with a large number of positive corresponding points and a small number of negative corresponding points, that is, the condition number with good performance.

Further, the matching result display process includes a process of displaying, on the display device 12, the corresponding score evaluation information of the identification image used for matching when the matching processing unit 262 of the image matching and identification device 26 executes the matching process. The corresponding score evaluation information is information based on the corresponding score calculated in the matching process, and may be, for example, a numerical value of the corresponding score as it is, or may be, for example, a plurality of levels or scores.

In this case, the matching result display processing section 333 displays the registered image display section 51, true value image display area 52, false value image display area 53, true value image on the screen of the display device 12, as shown in FIG. 23, for example. An evaluation information display area 541, a true value image correlation information display area 542, a false value image evaluation information display area 551, and a false value image correlation information display area 552 are displayed. The registered image display section 51 displays one of the registered images registered in the evaluation image registration section 242 of the recording device 24 among the images to be subjected to the matching process.

The true value image display area 52 displays identification images that are registered in the evaluation image registration section 242 of the recording device 24 and that are displayed on the registered image display section 51 among the images to be subjected to the matching process. Images of the same subject taken under different conditions are displayed. In the example of FIG. 23, the subject S1 of the image displayed in the registered image display section 51 and the subject S1 of the image displayed in the true value image display area 52 are the same individual. In this embodiment, an image obtained by photographing the same subject as the subject of the image displayed on the registered image display section 51 under different conditions is referred to as a true value image. In this case, the registered image and the identification image taken of the same subject are registered in the evaluation image registration unit 242 in a mutually linked state, that is, in a paired state. In other words, the matching result display processing unit 333 can recognize whether the images are of the same subject or different subjects by identifying the association between the registered image and the identification image.

The false value image display area 53 displays identification images that are registered in the evaluation image registration section 242 of the recording device 24 and are displayed on the registered image display section 51 among the images to be subjected to the matching process. Images taken of subjects different from the subject of the image under different conditions are displayed. In the example of FIG. 23, the subject S1 of the image displayed on the registered image display section 51 and the subject S2 of the image displayed on the true value image display area 52 are different individuals. In this embodiment, an image obtained by photographing a subject different from the subject of the image displayed on the registered image display section 51 is referred to as a false value image.

The true value image evaluation information display area 541 also displays the result of the comparison process between the image displayed in the registered image display section 51 and the image displayed in the true value image display area 52, that is, the true value. Corresponding score evaluation information for the image is displayed. In addition, the false value image evaluation information display area 551 displays the result of the comparison process between the image displayed in the registered image display section 51 and the image displayed in the false value image display area 53, that is, the false value. Corresponding score evaluation information for the image is displayed.

The matching result display process includes a process of displaying information based on the global feature correlation coefficient on the display device 12. In this case, the true value image correlation information display area 542 displays the global feature amount calculated from the registered image displayed on the registered image display section 51 and the value of the true value image displayed on the true value image display area 52. The correlation coefficient with the calculated global feature amount is displayed. The false value image correlation information display area 552 displays the global feature amount calculated from the registered image displayed on the registered image display section 51 and the global feature amount calculated from the false value image displayed on the false value image display area 53. The correlation coefficient obtained by comparing with the global feature amount is displayed.

For example, the user operates the input device 11 to perform the matching process on the images to be displayed in the registered image display section 51, true value image display area 52, false value image display area 53, and true value image evaluation information display area 541. Select an image and perform matching processing on that image. When the matching process is performed, the matching result display processing unit 333 displays the corresponding score evaluation information in the true value image evaluation information display area 541 and the false value image evaluation information display area 551 as the matching result, and The global feature correlation coefficients are displayed in the true value image correlation information display area 542 and the false value image correlation information display area 552, respectively. Thereby, the user can view the corresponding score evaluation information displayed in the true value image evaluation information display area 541 and the false value image evaluation information display area 551, as well as the corresponding score evaluation information displayed in the true value image correlation information display area 542 and the false value image correlation information display area 552. By looking at the information based on the global feature correlation coefficients displayed in , it is possible to check the performance of various conditions and parameters used in the matching process. In this case, the information based on the global feature correlation coefficient may be the value of the global feature correlation coefficient itself, or it may be the level or rank that processes the global feature correlation coefficient to express performance in several stages. It's good as well.

The matching result display processing unit 333 may cause the display device 12 to display the number of positive correspondence points or the number of negative correspondence points between the registered image and the identification image, as shown in FIG. 18, for example. Further, the matching result display processing unit 333 may separate and display the feature points that each individual image has uniquely and the feature points that they have in common, for example, between the registered image and the identification image. good. In this case, the matching result display processing unit 333 may display the number and proportion of feature points unique to each image and feature points common to each image. The matching result display processing unit 333 then selects a mode in which only the registered image and the identified image are displayed, a mode in which feature points are added to the registered image and the identified image, and a mode in which feature points and corresponding points are displayed in the registered image and the identified image. In addition, the display mode may be arbitrarily switched.

Here, for parts manufactured in the same process, such as industrial products, differences in characteristics between individuals are extremely small compared to biometric authentication such as a human face or fingerprint. Therefore, it is necessary to match a large number of features that appear on the surface of an individual, which requires a large amount of calculation and takes time to perform identification.The amount of calculation increases explosively as the population of objects to be identified increases.

On the other hand, the individual identification system 1 includes an image acquisition processing section 212, a feature point extraction processing section 221, a local feature calculation processing section 222, a local feature group classification processing section 223, a search target image registration processing section 231, a global feature It includes an amount registration processing section 234, a narrowing down processing section 2611, and a specific processing section 2612. The image acquisition processing unit 212 is capable of executing an image acquisition process of acquiring an image of a subject photographed by the photographing device 10. The feature point extraction processing unit 221 can execute a feature point extraction process of extracting feature points from the image acquired in the image acquisition process. The local feature amount calculation processing unit 222 is capable of executing a local feature amount calculation process that calculates the local feature amount of the feature point extracted in the feature point extraction process.

The local feature group classification processing unit 223 classifies the plurality of local features obtained by the local feature calculation process into a predetermined number, for example, 64 local feature groups, according to the value of the local feature. It is possible to perform group classification processing. The global feature amount calculation processing unit 224 is capable of executing global feature amount calculation processing that calculates global feature amounts based on each local feature amount group. The search target image registration processing unit 231 is capable of executing a search target image registration process of registering a plurality of search target images as registered images in the search target image registration unit 241 in advance. The global feature registration processing unit 234 is capable of executing a global feature registration process of registering the global feature related to the registered image in the global feature registration unit 244. For example, as shown in FIG. 15, the narrowing-down processing unit 2611 performs narrowing-down processing of selecting and narrowing down a predetermined number of registered images having global features highly correlated with the global features of the identified image from among a plurality of registered images. is possible. Then, as shown in FIG. 16, for example, the identification processing unit 2612 compares the local features of the candidate registered image extracted by the narrowing down process with the local features of the identified image, and determines the number of corresponding points of the local features. It is possible to perform identification processing to identify the most frequently registered images.

According to this configuration, the narrowing down processing unit 2611 roughly narrows down candidate registered images by comparing the global feature amounts of the registered image and the identification image by executing the narrowing down process. Although this narrowing down process requires less calculation and can be processed much faster than a specific process that compares all the local feature amounts of each image, it is difficult to achieve accuracy. Next, the identification processing unit 2612 performs identification processing to identify the target by comparing the registered images of the narrowed down candidates with the local feature amount of the identification image. Although this identification processing requires more calculations and takes more time than narrowing down processing that compares global features, it has good accuracy. In this way, the individual identification system 1 achieves high-speed and highly accurate retrieval by executing the two-step process of narrowing down using global features and specifying using local features. be able to. This is particularly effective when the number of registered images increases.

The individual identification system 1 further includes at least one of a photographing equipment condition adjustment processing section 311, a photographing target condition adjustment processing section 312, and a feature extraction condition adjustment processing section 321. The imaging equipment condition adjustment processing unit 311 is capable of executing imaging equipment condition adjustment processing for adjusting imaging equipment conditions, which are conditions related to the imaging equipment 10, based on user operations. The photographing target condition adjustment processing unit 312 is capable of executing photographing target condition adjustment processing for adjusting photographing target conditions, which are conditions related to the photographing target, based on a user's operation. Then, the feature extraction condition adjustment processing unit 321 can execute a feature extraction condition adjustment process that adjusts the feature extraction conditions that are conditions related to the feature point extraction process based on the user's operation.

According to this, the user operates at least one of the imaging equipment condition adjustment processing section 311, the imaging target condition adjustment processing section 312, and the feature extraction condition adjustment processing section 321 to set various conditions and parameters, that is, perform tuning. It can be carried out. As a result, even if the surface pattern changes during the manufacturing process, for example, when the surface pattern changes between the time of image registration and the time of identification, it can be handled flexibly, at high speed, and with precision. Can be well identified.

The individual identification system 1 includes a global feature dictionary generation processing section 251. The global feature dictionary generation processing unit 251 performs global feature dictionary generation processing that generates a global feature dictionary whose dictionary information includes representative values of each local feature group obtained from a plurality of training images acquired in advance. It is doable. The global feature amount calculation process includes processing to calculate a histogram obtained by voting on dictionary information having a high correlation with each local feature amount as a global feature amount. In addition, the global feature calculation process calculates the sum of the difference vectors between each local feature and the representative value, which is dictionary information with a high correlation, for the number of representative values, and then calculates the vector that connects them as the global feature. It may also include processing to do so.

In other words, the locations extracted as feature points vary depending on the type and condition of the subject, so it is not enough to simply classify the local features obtained from the image and calculate the global features. , the performance of global features may be low, that is, the discrimination ability may be low. In contrast, according to this configuration, by using a global feature dictionary that has dictionary information suitable for the subject, it is possible to obtain global features with good performance even if the type or state of the subject changes. As a result, the precision of the narrowing down process can be improved.

Furthermore, in this configuration, the global feature dictionary generation process includes a process of setting the centroid of the local feature group as a representative value in the dictionary information. The centroid of a local feature group is representative of the local feature group. Therefore, by setting the centroid of the local feature group as a representative value in the dictionary information, a high-performance global feature dictionary can be obtained.

The individual identification system 1 includes an image acquisition processing section 212 and a feature point extraction processing section 221, and further includes a feature point display processing section 331. The feature point display processing unit 331 is capable of executing a feature point display process in which feature points extracted by the feature point extraction process are displayed on the display device 12 by superimposing them on an image, as shown in FIG. 21, for example.

The individual identification system 1 includes at least one of a photographing equipment condition adjustment processing section 311, a photographing target condition adjustment processing section 312, and a feature extraction condition adjustment processing section 321. The imaging equipment condition adjustment processing unit 311 is capable of executing imaging equipment condition adjustment processing for adjusting imaging equipment conditions, which are conditions related to the imaging equipment 10, based on user operations. The photographing target condition adjustment processing unit 312 is capable of executing photographing target condition adjustment processing for adjusting photographing target conditions, which are conditions related to the photographing target, based on a user's operation. Then, the feature extraction condition adjustment processing unit 321 can execute a feature extraction condition adjustment process that adjusts the feature extraction conditions that are conditions related to the feature point extraction process based on the user's operation.

For example, when the individual identification system 1 is used at a production site, the parts to be identified may change in color, etc. of the surface pattern due to aging or heat processing during the manufacturing process. For parts that undergo processing that changes the color of the surface pattern during the manufacturing process, the positions and numbers of feature points extracted before and after the processing, that is, at the time of image registration and identification, and local feature amounts may change significantly, resulting in a decrease in identification accuracy.

On the other hand, according to the present configuration, the feature points extracted in the feature point extraction process are displayed on the display device 12 while being superimposed on the image. You can check to see if it is there. Then, the user selects the feature extraction condition adjustment processing unit 321, the imaging target condition adjustment processing unit 312, and At least one of the feature extraction condition adjustment processing units 321 can be operated to set various conditions and parameters, that is, to perform tuning, while checking the positions of feature points to be extracted. As a result, even if the surface pattern is processed to change during the manufacturing process, that is, the surface pattern changes between the time of image registration and the time of identification, it can be identified with high accuracy.

The individual identification system 1 includes a local feature amount calculation processing section 222 and a local feature amount group classification processing section 223. Then, as shown in FIG. 21(A), the feature point display process displays the feature points in different ways for each local feature group to which the local features belong, for example, in different colors, sizes, and shapes. 12 is further included.

According to this, the user can understand at a glance where and what kind of feature points are present in the image. This makes it easier to set various conditions and parameters while checking the positions of extracted feature points. As a result, it is possible to more accurately identify products that have been subjected to processing that changes the surface pattern during the manufacturing process.

The individual identification system 1 includes a global feature calculation processing section 224 and further includes a global feature calculation statistical information display processing section 332. As shown in FIG. 21(B), the global feature calculation statistical information display processing unit 332 displays the global It is possible to execute a global feature calculation statistical information display process that displays the feature calculation statistical information G on the display device 12.

According to this, the user can easily understand what kind of global feature amount has been obtained. That is, by displaying information on how many votes have been given to which dictionary information in the area feature dictionary, the user can grasp the bias in how the dictionary information included in the area feature dictionary is selected. This makes it easier to set various conditions and parameters regarding the global feature while looking at the displayed global feature calculation statistical information G. As a result, it is possible to more accurately identify products that have been subjected to processing that changes the surface pattern during the manufacturing process.

The individual identification system 1 further includes a global feature dictionary generation processing section 251. The global feature dictionary generation processing unit 251 performs global feature dictionary generation processing that generates a global feature dictionary whose dictionary information includes representative values of each local feature group obtained from a plurality of training images acquired in advance. It is doable. Then, as shown in FIG. 13, in the global feature calculation statistical information display process, the histogram or correlation value statistics obtained by voting for the dictionary information d that has a high correlation with each local feature in the global feature calculation process. It includes a process of displaying the score on the display device 12 as global feature value calculation statistical information G as shown in FIG.

According to this, the user can easily grasp the global feature amount visually and intuitively. This makes it easier to set various conditions and parameters regarding the global feature while looking at the displayed global feature calculation statistical information G. As a result, it is possible to more accurately identify products that have been subjected to processing that changes the surface pattern during the manufacturing process.

The individual identification system 1 further includes a global feature dictionary setting processing section 322. The global feature dictionary setting processing unit 322 calculates a score for each piece of global feature calculation statistical information, and automatically sets the one with the highest score as the global feature dictionary to be used in the local feature group classification process. Dictionary setting processing can be executed.

According to this, since an appropriate global feature dictionary is automatically set, the user does not need to manually adjust the conditions and parameters one by one, and as a result, user convenience is improved.

When the individual identification system 1 having this configuration is used, for example, in a factory or the like, it is assumed that the registration image and the identification image are taken at different locations. In this case, dust generated in the factory, lighting inside the factory, natural light such as outside light, and surrounding noise sources become disturbance factors, and the shooting environment between the registration image and the identification image differs greatly. there is a possibility. Then, the extracted feature points, local features, etc. may change, and the identification performance may deteriorate.

Therefore, the individual identification system 1 with this configuration includes an image acquisition processing section 212, a feature point extraction processing section 221, a local feature calculation processing section 222, a photographing target condition adjustment processing section 312, and a photographing equipment condition adjustment processing section. 311. According to this, the user operates the imaging target condition adjustment processing unit 312 and the imaging equipment condition adjustment processing unit 311 to set the imaging target conditions and imaging equipment conditions suitable for extracting feature points and calculating local feature amounts. Can be adjusted. Therefore, by setting appropriate conditions according to changes in the environment, it is possible to reduce the influence of external disturbances, and it is possible to appropriately highlight the desired characteristic pattern. As a result, identification accuracy can be improved.

The individual identification device 20 further includes a brute force condition generation processing section 313. The brute force condition generation processing unit 313 is capable of executing brute force condition generation processing in which a combination of photographing target conditions within a preset range and photographing equipment conditions within a preset range is generated by brute force. The image acquisition process includes a process of automatically acquiring images of all the conditions generated in the round-robin condition generation process.

According to this, it is possible to automatically photograph and acquire a large number of images with different photographing target conditions and photographing equipment conditions. Therefore, when searching for an optimal combination of conditions by changing the photographing target conditions and photographing equipment conditions, the user does not have to manually generate the combination of conditions and perform photographing. As a result, it is possible to reduce the effort required to adjust the photographing target conditions and the photographing equipment conditions. Furthermore, since conditions with good performance can be easily set, identification performance can also be improved.

The individual identification device 20 further includes an optimal condition presentation processing section 314. The optimal condition presentation processing unit 314 determines whether, among the registered images and identification images taken under all the conditions generated in the round-robin condition generation process, the number of correspondences between the registered image and the identification image taken of the same subject is large. It is possible to execute optimal condition presentation processing that presents a combination of photographing target conditions and photographing equipment conditions as optimal conditions. According to this, the user can easily obtain and set the optimal conditions with a large number of feature points, so that it is possible to further reduce the effort of adjusting the conditions of the photographing object and the conditions of the photographing equipment. Furthermore, since the optimum combination of the photographing object conditions and the photographing equipment conditions can be easily set, the identification performance can also be improved.

In the optimal condition presentation process, multiple feature extraction conditions, which are conditions related to the feature point extraction process, are prepared in advance, and registered images and identification images taken of the same subject under all conditions generated in the round-robin condition generation process are prepared in advance. This includes a process of presenting a combination of conditions that will result in the largest number of corresponding points when compared using the feature points obtained by executing the feature point extraction process under a plurality of feature extraction conditions. According to this, the user can easily obtain the optimal conditions among the combinations of the imaging target conditions and the imaging equipment conditions, as well as the feature extraction conditions related to the feature point extraction process. Thereby, it is possible to further reduce the effort required to adjust the imaging target conditions, imaging equipment conditions, and feature extraction conditions. Since the optimal combination of the imaging target conditions, imaging equipment conditions, and feature extraction conditions can be easily set, the identification performance can be further improved.

For example, unlike human faces or fingerprints, the characteristics of the patterns that appear on the parts of industrial products vary depending on the type, material, and processing method of the object to be compared and identified. Therefore, it is necessary to adjust the parameters of the algorithm used for verification and identification to appropriate values depending on the type of object to be identified. However, there are many parameter items, and it takes a lot of effort and time for the user to manually search for the optimal combination among all the combinations.

Therefore, the individual identification system 1 includes a feature point extraction processing section 221 and a local feature calculation processing section 222. The individual identification system 1 further includes a matching processing section 262, a parameter set generation processing section 324, and a parameter set setting processing section 325. The matching processing unit 262 can perform a matching process of comparing the local feature amounts of the registered image and the identified image, obtaining the number of corresponding feature points, and comparing the registered image and the identified image. The parameter set generation processing unit 324 generates a plurality of parameter sets by combining each parameter set to a setting value within a setting range defined for each parameter for parameters for setting conditions for matching processing. Generation processing can be executed. The parameter set setting processing unit 325 can execute a parameter set setting process of setting a parameter set based on a user's operation.

The matching process includes a process of executing the matching process using each parameter set generated in the parameter set generation process. The individual identification system 1 further includes a matching result display processing section 333. The matching result display processing unit 333 is capable of executing matching result display processing in which matching results obtained when matching processing is executed using each parameter set are collectively displayed on the display device 12.

According to this, a parameter set that is a combination of parameters is automatically generated, and a matching process is performed using each of the parameter sets. Therefore, the user does not need to manually create and check each combination of parameters. The comparison results are displayed all together on the display device 12. Therefore, the user can check the performance of each parameter set by viewing the results displayed on the display device 12. Then, the user can view the results displayed on the display device 12 and select an appropriate parameter set. Thereby, regardless of the type of target to be verified and identified, the parameters of the algorithm used for verification and identification can be easily and easily and accurately adjusted.

As shown in FIG. 22, the matching result display process includes a process of displaying a graph of the matching results obtained by performing the matching process using each parameter set. It becomes easier to visually and intuitively understand which parameter set has the best performance among multiple parameter sets. This makes it possible to easily adjust parameters with high accuracy.

The individual identification system 1 further includes an evaluation image registration processing section 232. The evaluation image registration processing unit 232 generates a registered image group including a plurality of registered images each photographed of a different subject, an identification image of the same subject as one of the registered image groups but photographed under different conditions, It is possible to execute an evaluation image registration process in which the images are registered in the evaluation image registration unit 242 in a state where they are mutually linked as images for evaluation in the matching process.

According to this, the user does not need to take images each time when evaluating a parameter set. As a result, parameter sets can be easily evaluated. Further, according to this, registered images and identification images taken of the same subject are registered in a mutually linked state. Therefore, when performing matching processing using registered images and identification images, it is not clear in advance the correspondence between the registered images and identification images, that is, whether they were taken of the same subject or different subjects. ing. As a result, by using images with clear correspondence relationships, it is possible to clarify the desired result in the matching process, and it becomes easier for the user to evaluate the parameter set used in the matching process.

In other words, when the matching process is performed using registered images and identification images taken of the same subject, the number of corresponding points is large, and the matching process is performed using registered images and identification images taken of different subjects. If the number of corresponding points is small in a case, the user can evaluate that the performance of the parameter set used for the matching process is high. Conversely, if the matching process is performed using registered images and identification images taken of the same subject, the number of corresponding points is small, or the matching process is performed using registered images and identification images taken of different subjects. In this case, if the number of corresponding points is small, the user can evaluate that the performance of the parameter set used for the matching process is low.

The matching process includes a process of matching one or more registered images included in the registered image group with the identification image. As shown in FIG. 23, the matching result display process includes a process of displaying corresponding score evaluation information on the display device 12 based on the corresponding scores calculated in the matching process. According to this, by viewing the corresponding score evaluation information displayed on the display device 12, the user can easily evaluate the performance of the parameter set, and thereby select appropriate parameters according to the type of target to be collated and identified. You can choose a set. As a result, high verification and identification performance can be achieved for any type of target.

The individual identification device 20 further includes a feature extraction condition adjustment processing section 321. The feature extraction condition adjustment processing unit 321 can execute feature extraction condition adjustment processing. The feature extraction condition adjustment process calculates the correct correspondence score, which is the number of correspondence points between the registered image and the identified image, which are images of the same subject as the identified image in the registered image group, and the difference between the registered image group and the identified image. It includes a process of calculating a negative correspondence score, which is the correspondence score between the registered image of the subject and the identification image. The feature extraction condition adjustment process includes a process of adjusting the feature extraction conditions based on the local feature amount calculation process so that the ratio between the number of positive corresponding points and the number of negative corresponding points becomes the largest.

In this case, the feature extraction condition adjustment processing unit 321 searches for a feature extraction condition with a large number of positive corresponding points and a small number of negative corresponding points so that the ratio between the number of positive corresponding points and the number of negative corresponding points is the largest. A large number of corresponding points means that there are many corresponding points between images taken of the same subject. This means that in the matching process between images that depict the same subject, it is determined that the subjects included in the images are likely to be the same. It means high performance. Furthermore, a small number of negative corresponding points means that there are few corresponding points between images taken of different subjects. This means that in the matching process between images depicting different subjects, it is identified that the subjects included in the images are likely to be different; in other words, the discrimination performance between images depicting different subjects is high. means.

In this way, the feature extraction condition adjustment processing unit 321 automatically adjusts the feature extraction conditions so that the matching and discrimination performance between images depicting the same subject and between images depicting different subjects is improved. It is possible to save the user's effort in making adjustments while obtaining high matching and identification performance.

The individual identification system 1 includes a local feature group classification processing section 223 and a global feature calculation processing section 224. The individual identification system 1 further includes a global feature correlation coefficient calculation processing section 323. The global feature correlation coefficient calculation processing unit 323 is a global feature correlation coefficient that compares the global feature of each registered image with the global feature of the identified image and calculates a global feature correlation coefficient that indicates the correlation. It is possible to perform calculation processing. The matching result display process includes a process of displaying information based on the global feature correlation coefficient on the display device 12.

The global feature correlation coefficient is one index that indicates the degree of matching between the global feature amounts between the registered image and the identified image, that is, the performance. By displaying information based on the global feature correlation coefficient on the display device 12, the user can easily understand the performance of the global feature. Therefore, the user can adjust various conditions and parameters using the information based on the global feature correlation coefficient on the display device 12 as a guide, which makes the user's work easier.

The individual identification system 1 further includes a global feature dictionary registration section 245 and a global feature dictionary setting processing section 322. The global feature dictionary setting processing unit 322 can execute global feature dictionary setting processing. The global feature dictionary setting process selects a registered image captured of the same subject and an identification image based on a global feature correlation coefficient from among a plurality of global feature dictionaries registered in the global feature dictionary registration unit 245. It includes a process of selecting a global feature dictionary that can be determined to be the same, and determining that a registered image and an identification image obtained by photographing different subjects are not the same, and setting it as the global feature dictionary used in the local feature group classification process.

As described above, the global feature correlation coefficient serves as an index indicating the degree of matching of the correlation between the global feature amounts in the registered image and the identification image. The greater the correlation between the global feature amounts between two images, the higher the possibility that the same subject is included in both images. Therefore, the fact that the correlation between the global feature amounts between images taken of the same subject is high means that the degree of coincidence between the global feature amounts calculated for the two images taken of the same subject is high. In other words, this means that the performance of the global feature dictionary is high among images taken of the same subject.

On the other hand, the smaller the correlation between the global feature amounts between two images, the higher the possibility that the subjects included in the two images are different. Therefore, a low correlation of global features between images of different subjects means that the degree of agreement between the global features calculated using the global feature dictionary for images of different subjects is low. do. In other words, this means that the performance of the global feature dictionary is high between images taken of different subjects.

According to this configuration, the global feature dictionary setting process automatically sets a global feature dictionary that has high performance between images depicting the same subject and also has high performance between images depicting different subjects. be able to. Therefore, the user's effort for adjustment can be saved, and high verification and identification performance can be obtained.

In the global feature dictionary setting process, when two global features are vectors, the absolute value of each is normalized to 1 and the inner product is taken, and if the value of the inner product is close to 1, the identity is high. It may also include a process of determining that the identity is low if the value of the inner product is close to -1. Furthermore, the global feature correlation coefficient can be a value of the distance between vectors of two global features. In this case, the global feature dictionary setting process includes a process in which it is determined that the identity is high if the distance between the vectors of the two global features is close to 0, and that the identity is low if it is large. According to these configurations, the performance of the global feature dictionary can be quantitatively evaluated.

The individual identification system 1 includes an individual identification device 20 having an image acquisition processing section 212, a feature point extraction processing section 221, and a specific processing section 2612. In addition to the individual identification device 20, the individual identification system 1 includes at least two of a visualization device 33, a photographing condition adjustment device 31, and a verification identification condition adjustment device 32.

The visualization device 33 includes a feature point display processing section 331. The feature point display processing unit 331 can execute a feature point display process in which the feature points extracted by the feature point extraction process are displayed on the display device 12 in a superimposed manner on the image.

The photographing condition adjustment device 31 compares the local feature amounts of a registered image photographing the same subject and an identified image among a plurality of registered images registered in advance, obtains the number of corresponding feature points, and then adjusts the registered image and the identified image. It causes the individual identification device 20 to execute a matching process for comparing the images, and includes a photographing target condition adjustment processing section 312 and a photographing condition adjustment device 31. The photographing target condition adjustment processing unit 312 is capable of executing photographing target condition adjustment processing for adjusting photographing target conditions, which are conditions related to the photographing target. The photographing equipment condition adjustment processing unit 311 is capable of executing photographing equipment condition adjustment processing for adjusting photographing equipment conditions that are conditions related to the photographing equipment 10 .

The matching identification condition adjustment device 32 includes a parameter set generation processing section 324 and a parameter set setting processing section 325. The parameter set generation processing unit 324 generates a plurality of parameter sets by combining each parameter set to a set value within a predetermined range for each parameter for parameters for setting conditions for matching processing. Processing is possible. The parameter set setting processing unit 325 can execute a parameter set setting process of setting a parameter set based on a user's operation. Then, the matching identification condition adjustment device 32 causes the matching processing unit 262 to execute a process of executing a matching process using each parameter set generated in the parameter set generation process. The matching identification condition adjustment device 32 further includes a matching result display processing section 333 capable of executing a matching result display process of collectively displaying the matching results of the matching process on the display device 12 using the parameter set.

Since the individual identification system 1 includes the visualization device 33, the feature points extracted in the feature point extraction process are superimposed on the image and displayed on the display device 12, so the user can determine which location in the image is the feature point. You can check to see if it has been extracted. Then, the user sets various conditions and parameters while checking the positions of the extracted feature points, that is, performs tuning, so that feature points can be extracted focused on areas where the surface pattern is difficult to change during the manufacturing process. I can do it. As a result, it is possible to identify items with high accuracy even if the surface pattern changes during the manufacturing process, that is, even if the surface pattern changes between the time of image registration and the time of identification. Excellent effects can be obtained.

Since the individual identification system 1 includes the imaging condition adjustment device 31, the user can operate the imaging target condition adjustment processing unit 312 and the imaging equipment condition adjustment processing unit 311 to extract feature points and calculate local feature amounts. It is possible to adjust suitable photographing target conditions and photographing equipment conditions. Therefore, by setting appropriate conditions according to changes in the environment, the influence of disturbances can be reduced, and the characteristic pattern can be appropriately highlighted. As a result, excellent effects such as improved identification accuracy can be obtained.

Since the individual identification system 1 includes the comparison identification condition adjustment device 32, the user does not need to manually create each combination of parameters and perform the comparison. The comparison results are displayed all together on the display device 12. Therefore, the user can check the performance of each parameter set by viewing the results displayed on the display device 12. Then, the user can view the results displayed on the display device 12 and select an appropriate parameter set. This provides an excellent effect in that the parameters of the algorithm used for verification and identification can be easily and easily and accurately adjusted regardless of the type of object to be verified and identified.

The individual identification system 1 simultaneously provides at least two of the visualization device 33, the photographing condition adjustment device 31, and the collation identification condition adjustment device 32, thereby simultaneously achieving two or more of the excellent effects of the above-mentioned devices. Obtainable. As a result, it is possible to respond to changes in the surface pattern of the identification target, to make it less susceptible to external disturbances, to reduce the user effort required to adjust conditions and parameters, and to improve the precision of accuracy conditions and parameters. As a result, the performance of verification and identification can be improved.

The individual identification system 1 includes a visualization device 33, a photographing condition adjustment device 31, and a comparison identification condition adjustment device 32. According to this, by providing the visualization device 33, the photographing condition adjustment device 31, and the comparison identification condition adjustment device 32 all at the same time, all the excellent effects of the above-mentioned devices can be obtained at the same time. That is, according to this individual identification system 1, it is possible to respond to changes in the surface shape of the identification target, to perform adjustments that are less susceptible to external disturbances, and to further reduce the user's effort in adjusting conditions and parameters. I can do it.

The search processing unit 261 further includes a narrowing down processing unit 2611 that can execute narrowing down processing. The narrowing process selects a predetermined number of registered images having global features that are highly correlated with the global features of the identified image from among the plurality of registered images registered in the search target image registration unit 241 as candidates for the identification process. Includes narrowing down processing.

According to this, the search process can be performed at high speed by performing two stages of narrowing down processing using global feature amounts and specifying processing using local feature amounts. As a result, according to the individual identification system 1 with this configuration, it is possible to achieve high matching identification performance and high-speed retrieval.

Although the present disclosure has been described based on examples, it is understood that the present disclosure is not limited to the examples or structures. The present disclosure also includes various modifications and equivalent modifications. In addition, various combinations and configurations, as well as other combinations and configurations that include only one, more, or fewer elements, are within the scope and scope of the present disclosure.

The control unit and the method described in the present disclosure are implemented by a dedicated computer provided by configuring a processor and memory programmed to perform one or more functions embodied by a computer program. may be done. Alternatively, the controller and techniques described in this disclosure may be implemented by a dedicated computer provided by a processor configured with one or more dedicated hardware logic circuits. Alternatively, the control unit and the method described in the present disclosure may be implemented by a combination of a processor and memory programmed to perform one or more functions and a processor configured by one or more hardware logic circuits. It may be implemented by one or more dedicated computers configured. The computer program may also be stored as instructions executed by a computer on a computer-readable non-transitory tangible storage medium.

1... Individual identification system, 10, 10a, 10b... Imaging equipment, 212... Image acquisition processing section, 221... Feature point extraction processing section, 222... Local feature calculation processing section, 311... Imaging equipment condition adjustment processing section, 312... Shooting target condition adjustment processing section, 313... Round-robin condition generation processing section, 314... Optimum condition presentation processing section

Claims (2)

an image acquisition processing unit (212) capable of executing an image acquisition process of acquiring images of the identification target taken by the registration-side and identification-side imaging devices (10a, 10b) as registration images and identification images, respectively;
a feature point extraction processing unit (221) capable of executing a feature point extraction process for extracting feature points included in the registered image and the identification image acquired in the image acquisition process;
a local feature calculation processing unit (222) capable of executing a local feature calculation process that calculates a local feature of the feature point extracted in the feature point extraction process;
a photographic subject condition adjustment processing unit (312) capable of executing a photographic subject condition adjustment process that adjusts photographic subject conditions that are conditions related to the photographic subject;
a photographic equipment condition adjustment processing unit (311) capable of executing a photographic equipment condition adjustment process that adjusts photographic equipment conditions that are conditions related to the photographic equipment ;
a brute force condition generation processing unit (313) capable of executing a brute force condition generation process of generating a combination of the photographing target condition within a preset range and the photographing equipment condition within a preset range by brute force; and ,
The image acquisition process includes a process of acquiring the registered image and the identification image taken under all conditions generated in the round-robin condition generation process,
The registered images and the identification images taken under all the conditions generated in the round-robin condition generation process.
A pair obtained by comparing the registered image taken of the same subject and the identified image among different images.
A combination of the imaging target condition and the imaging equipment condition with a large number of response points is presented as the optimal condition.
further comprising an optimum condition presentation processing unit (314) capable of executing an optimum condition presentation process to
Individual identification system (1). In the optimal condition presentation process, a plurality of feature extraction conditions, which are conditions related to the feature point extraction process, are prepared in advance, and the registered image and the including a process of presenting a combination of conditions that will result in the largest number of corresponding points when matched using feature points obtained by executing the feature point extraction process under a plurality of feature extraction conditions prepared in advance for the identification image;
The individual identification system according to claim 1 .

Images