Information processing device, information processing method, and recording medium

Abstract

An information processing device according to the present invention comprises: an acquisition means that acquires a plurality of first processed images which are obtained by subjecting first pattern images to respective first image processes and a plurality of second processed images which are obtained by subjecting second pattern images to respective second image processes; and an output means that outputs a plurality of search target comparison pairs for use in a search process, among comparison pairs in each of which one of the plurality of first processed images is compared with one of a plurality of second processed images. The search process is a process for searching for a combination of a comparison pair. The plurality of search target comparison pairs are specified such that comparison characteristics in the plurality of search target comparison pairs are not similar to one another.

Description

Information processing device, information processing method, and recording medium

[0001] The present disclosure relates to the technical fields of an information processing device, an information processing method, and a recording medium.

[0002] Patent Document 1 describes that a genetic algorithm or the like can be used as a method for optimizing control parameters, and that by reducing the search range of the control parameters, the calculation time required for optimizing the control parameters can be shortened.

[0003] Japanese Patent Application Laid-Open No. 2017-157112

[0004] An object of this disclosure is to provide an information processing device, an information processing method, and a recording medium that aim to improve upon the techniques disclosed in prior art documents.

[0005] One aspect of the information processing device includes an acquisition means for acquiring a plurality of first processed images obtained by applying a different first image processing to a first pattern image, and a plurality of second processed images obtained by applying a different second image processing to a second pattern image, and an output means for outputting a plurality of search-target match pairs to be used in a search process from among match pairs that match any one of the plurality of first processed images with any one of the plurality of second processed images, wherein the search process is a process for searching for combinations of the match pairs, and the plurality of search-target match pairs are specified so that the match characteristics of each of the plurality of search-target match pairs are not similar.

[0006] One aspect of the information processing method includes obtaining a plurality of first processed images obtained by performing a first image process that is different for each of a first pattern image, and a plurality of second processed images obtained by performing a second image process that is different for each of a second pattern image, and outputting a plurality of search-target match pairs to be used in a search process from among match pairs that match any one of the plurality of first processed images with any one of the plurality of second processed images, the search process being a process of searching for combinations of the match pairs, and the plurality of search-target match pairs being specified so that the match characteristics of each of the plurality of search-target match pairs are not similar.

[0007] In one aspect of the recording medium, a computer program is recorded to cause a computer to execute an information processing method, which includes acquiring a plurality of first processed images obtained by applying a different first image processing to a first pattern image and a plurality of second processed images obtained by applying a different second image processing to a second pattern image, and outputting a plurality of search-target match pairs to be used in a search process from among match pairs that match one of the plurality of first-processed images with one of the plurality of second-processed images, the search process being a process of searching for combinations of the match pairs, and specifying the plurality of search-target match pairs so that the match characteristics of each of the plurality of search-target match pairs are not similar.

[0008] FIG. 1 is a block diagram showing an example of the configuration of an information processing device according to the present disclosure. FIG. 2 is a block diagram showing an example of the configuration of an information processing device according to the present disclosure. FIG. 3 is a flowchart showing an example of the information processing operation of an information processing device according to the present disclosure. FIG. 4 is an example of an image showing an example of the information processing operation of an information processing device according to the present disclosure. FIG. 5 is a block diagram showing an example of the configuration of an information processing device according to the present disclosure. FIG. 6 is a flowchart showing an example of the information processing operation of an information processing device according to the present disclosure. FIG. 7 is a block diagram showing an example of the configuration of an information processing device according to the present disclosure.

[0009] Hereinafter, embodiments of an information processing device, an information processing method, and a recording medium will be described with reference to the drawings. [1: First Embodiment]

[0010] A first embodiment of an information processing device, an information processing method, and a recording medium will be described below. Hereinafter, the first embodiment of the information processing device, the information processing method, and the recording medium will be described using an information processing device 1 according to this disclosure. [1-1: Configuration of Information Processing Device 1]

[0011] 1 is a block diagram showing the configuration of an information processing device 1 according to this disclosure. As shown in FIG. 1, the information processing device 1 includes an acquisition unit 11 and an output unit 12.

[0012] The acquisition unit 11 acquires a plurality of first processed images I1 and a plurality of second processed images I2. The plurality of first processed images I1 are images obtained by applying different first image processes to a first pattern image. The plurality of second processed images I2 are images obtained by applying different second image processes to a second pattern image.

[0013] The output unit 12 outputs a plurality of search target matching pairs (referred to as "search processing matching pairs") to be used in the search processing from among matching pairs (referred to as "matching pairs") that match one of the plurality of first processed images I1 with one of the plurality of second processed images I2. The plurality of search target matching pairs are specified so that the matching characteristics (referred to as "matching characteristics") of each of the plurality of search target matching pairs are not similar. The matching characteristics include at least one of properties possessed by the processed images used for matching and properties related to the matching result.

[0014] The search process is a process of searching for a combination of match pairs. Before performing the search process, the information processing device 1 reduces the number of match pairs to be searched. [1-2: Technical Effects of the Information Processing Device 1]

[0015] Even when matching is performed using different matching pairs, the matching characteristics may be similar. Even if matching pairs with similar matching characteristics are excluded from the search targets, the search range of the search process is not limited and the processing results are not adversely affected. In addition, the processing load of the search process can be reduced.

[0016] The information processing device 1 according to this disclosure excludes match pairs having similar match characteristics from the search process, thereby reducing the number of match pairs to be searched, thereby reducing the processing load of the search process. [2: Second Embodiment]

[0017] A second embodiment of an information processing device, an information processing method, and a recording medium will be described below. Hereinafter, the second embodiment of an information processing device, an information processing method, and a recording medium will be described using an information processing device 2 according to this disclosure. [2-1: Quality of pattern image]

[0018] In this disclosure, the first pattern image is a pattern image of lower quality than the second pattern image. For example, the first pattern image may include at least one of a pattern image with significant image distortion and a pattern image with a small area where the pattern is clear. In contrast, the second pattern image may include at least one of a pattern image with no significant image distortion and a pattern image with a large area where the pattern is clear.

[0019] In this disclosure, the second pattern image may be a pattern image from which more features usable for image matching can be extracted than the first pattern image. The second pattern image may be a pattern image from which more features useful for image matching can be extracted than the first pattern image. Also, in this disclosure, the first pattern image may be a pattern image from which a greater number of types of image processing can be performed when extracting features usable for matching than the second pattern image. The first pattern image may be a pattern image from which a greater number of types of image processing can be performed when extracting features useful for matching than the second pattern image. In other words, the first image processing may include more image processing than the second image processing. Also, the number of types of processing included in each of the image processing included in the first image processing may be greater than the number of types of processing included in each of the image processing included in the second image processing.

[0020] In this disclosure, the first pattern image may be an image of a latent fingerprint. A latent fingerprint is a fingerprint left at a crime scene or the like. Images of latent fingerprints often have significant image distortion and the area of ​​clear ridge regions is small. Furthermore, images of latent fingerprints are often images of a small region that captures only a portion of the entire fingerprint, and are of relatively low quality.

[0021] In this disclosure, the second pattern image may be an image of an imprint fingerprint. An imprint fingerprint is a fingerprint collected for the purpose of registering it in a database, for example. An image of an imprint fingerprint rarely has significant image distortion, has a large ridge area, and is of relatively high quality.

[0022] In fingerprint image matching, feature information extracted from one fingerprint image is compared with feature information extracted from another fingerprint image. The feature information includes information about feature points extracted from at least one of the end points and bifurcation points of fingerprint ridges. Based on the fingerprint image matching results, it is determined whether the fingerprint images were obtained from the same person.

[0023] When relatively high-quality fingerprint images, such as impressions, are used, fingerprint images acquired from the same person often have corresponding minutiae in similar positions.Furthermore, from relatively high-quality fingerprint images, such as impressions, it is often possible to extract a sufficient number of minutiae for matching.

[0024] In contrast, when a fingerprint image of relatively low quality, such as an image of a latent fingerprint, is used, there are often no feature points that correspond to similar positions even between fingerprint images obtained from the same person. Also, it is often not possible to extract a sufficient number of feature points for matching from a fingerprint image of relatively low quality, such as an image of a latent fingerprint. For this reason, when a fingerprint image of relatively high quality, such as an image of a latent fingerprint, is used, in order to extract features that are useful for matching, it is necessary to perform more image processing than is performed on an image of an inked fingerprint.

[0025] The feature information extracted from a fingerprint image varies depending on the image processing applied to the fingerprint image. Therefore, it is necessary to apply image processing that can extract feature information useful for matching. For example, even when the same image processing is applied, there are cases where features useful for matching can be extracted from an image of an inked fingerprint, while features useful for matching cannot be extracted from an image of a latent fingerprint. Similarly, even when the same image processing is applied, there are cases where features useful for matching cannot be extracted from an image of an inked fingerprint, while features useful for matching can be extracted from an image of a latent fingerprint.

[0026] Furthermore, among pairs of processed images that have been subjected to image processing (pairs of processed inked fingerprint images and processed latent fingerprint images: matching pairs), there are matching pairs that allow for accurate fingerprint matching and matching pairs that do not allow for accurate fingerprint matching. Therefore, in order to achieve high accuracy in the matching process between latent fingerprints and inked fingerprints, it is necessary to search for matching pairs between processed inked fingerprint images and processed latent fingerprint images. [2-2: Fusion Matching and Search Process]

[0027] Compared to the results of matching using a single matching pair, the results of fusion matching, which combines (fuse) the results of matching using multiple matching pairs, often achieve more accurate matching.Fusion matching also often improves matching accuracy by using many matching pairs.Fusion matching often becomes more advantageous the more variations there are in matching.

[0028] Each of the matching pairs has a different combination of the first image processing and the second image processing, and the number of matching pairs can be increased by increasing the number of the first image processing and the number of the second image processing.

[0029] The number of first image processing steps can be increased by dividing the first image processing into multiple processing steps and preparing multiple types of each processing step. For example, the first image processing may be an image processing that combines any one of 11 types of first image enhancement processing, any one of 12 types of residual processing, any one of 13 types of first image enhancement processing, and any one of 14 types of feature extraction processing. In this case, there are L (= 11 × 12 × 13 × 14) types of first image processing, so a maximum of L types of first processed images I1 can be prepared.

[0030] Similarly, the number of second image processing steps can be increased by dividing the second image processing into multiple processing steps and preparing multiple types of each processing step. For example, the second image processing step may be an image processing step that combines one of t1 types of image enhancement processing and one of t2 types of feature extraction processing. In this case, since there are T (= t1 × t2) types of second image processing steps, a maximum of T types of second processed images I2 can be prepared.

[0031] Furthermore, the accuracy of fusion verification varies depending on the combination of verification pairs. In other words, the magnitude of the effect of fusion verification varies depending on the combination of verification pairs. In other words, the maximum effect of fusion verification can be obtained by searching for the optimal combination of verification pairs. The search process in this embodiment is a process of searching for the optimal combination of verification pairs to maximize the effect of fusion verification.

[0032] To obtain the maximum effect of fusion matching, it is preferable that there are many candidate combinations (different matching pairs). On the other hand, when searching for a combination of matching pairs, even if there are multiple similar matching pairs, the effect of fusion matching cannot be increased. In other words, it is possible to search for an optimal combination without subjecting all of the matching pairs that can be prepared to the search process. Furthermore, the greater the number of matching pairs to be searched for, the heavier the processing load of the search process. Therefore, if all of the matching pairs that can be prepared are subject to the search process, the processing load of the search process will be heavy.

[0033] In particular, when matching pattern images, such as latent fingerprint images, which can be subjected to many types of image processing, the number of matching pairs that can be prepared is greater than when matching high-quality pattern images, such as impression fingerprint images. Therefore, when matching latent fingerprint images, there is a strong need to reduce the number of matching pairs and lighten the processing load of the search process.

[0034] Therefore, in this embodiment, the matching pairs to be used in the search process are identified before the search process, and useful preparations for the search process are made. [2-3: Configuration of the information processing device 2]

[0035] 2 is a block diagram showing the configuration of the information processing device 2. As shown in FIG. 2, the information processing device 2 includes a calculation device 21 and a storage device 22. The information processing device 2 may further include a communication device 23, an input device 24, and an output device 25. However, the information processing device 2 does not necessarily have to include at least one of the communication device 23, the input device 24, and the output device 25. The calculation device 21, the storage device 22, the communication device 23, the input device 24, and the output device 25 may be connected via a data bus 26.

[0036] The arithmetic device 21 includes, for example, at least one of a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), and an FPGA (Field Programmable Gate Array). The arithmetic device 21 reads a computer program. For example, the arithmetic device 21 may read a computer program stored in the storage device 22. For example, the arithmetic device 21 may read a computer program stored in a computer-readable, non-transitory recording medium using a recording medium reading device (e.g., an input device 24 described later) not shown in the drawings that is included in the information processing device 2. The arithmetic device 21 may acquire (i.e., download or read) the computer program from a device (not shown) located outside the information processing device 2 via the communication device 23 (or another communication device). The arithmetic device 21 executes the read computer program. As a result, logical functional blocks for executing operations to be performed by the information processing device 2 are realized within the arithmetic device 21. In other words, the arithmetic device 21 can function as a controller for realizing logical functional blocks for executing operations (in other words, processes) to be performed by the information processing device 2. The arithmetic device 21 may output information to devices (not shown), such as other computers or cloud servers, that are provided outside the information processing device 2, via the communication device 23 (or other communication devices).

[0037] FIG. 2 illustrates an example of logical functional blocks implemented within the computing device 21 to perform information processing operations. As illustrated in FIG. 2 , the computing device 21 includes an acquisition unit 211, which is a specific example of an "acquisition means" described in the appendix, an output unit 212, which is a specific example of an "output means" described in the appendix, a matching unit 213, which is a specific example of a "matching means" described in the appendix, a calculation unit 214, which is a specific example of a "calculation means" described in the appendix, and a conversion unit 215, which is a specific example of a "identification means" described in the appendix. However, at least one of the matching unit 213, the calculation unit 214, and the conversion unit 215 does not necessarily need to be implemented within the computing device 21. The calculation unit 214 may include a matching score correlation calculation unit 2141 and a distance matrix calculation unit 2142. Details of the operations of the acquisition unit 211, the output unit 212, the matching unit 213, the calculation unit 214, and the conversion unit 215 will be described later with reference to FIG. 3 .

[0038] The storage device 22 can store desired data. For example, the storage device 22 may temporarily store a computer program executed by the arithmetic device 21. The storage device 22 may temporarily store data that the arithmetic device 21 temporarily uses when the arithmetic device 21 is executing a computer program. The storage device 22 may store data that the information processing device 2 stores long-term. The storage device 22 may include at least one of a RAM (Random Access Memory), a ROM (Read Only Memory), a hard disk device, a magneto-optical disk device, an SSD (Solid State Drive), and a disk array device. In other words, the storage device 22 may include a non-temporary recording medium.

[0039] The communication device 23 is capable of communicating with devices external to the information processing device 2 via a communication network (not shown). The communication device 23 may be a communication interface based on standards such as Ethernet (registered trademark), Wi-Fi (registered trademark), Bluetooth (registered trademark), or USB (Universal Serial Bus).

[0040] The input device 24 is a device that accepts information input to the information processing device 2 from outside the information processing device 2. For example, the input device 24 may include an operation device (e.g., at least one of a keyboard, a mouse, and a touch panel) that can be operated by an operator of the information processing device 2. For example, the input device 24 may include a reading device that can read information recorded as data on a recording medium that can be externally attached to the information processing device 2.

[0041] The output device 25 is a device that outputs information to the outside of the information processing device 2. For example, the output device 25 may output information as an image. That is, the output device 25 may include a display device (a so-called display) that can display an image showing the information to be output. For example, the output device 25 may output information as sound. That is, the output device 25 may include an audio device (a so-called speaker) that can output sound. For example, the output device 25 may output information on paper. That is, the output device 25 may include a printing device (a so-called printer) that can print desired information on paper. [2-4: Information Processing Operation Performed by the Information Processing Device 2]

[0042] The information processing operation performed by the information processing device 2 will be described with reference to Fig. 3. Fig. 3 is a flowchart showing an example of the flow of the information processing operation performed by the information processing device 2.

[0043] As shown in FIG. 3, the acquisition unit 211 acquires a plurality of first processed images I1 and a plurality of second processed images I2 (step S20).

[0044] The multiple first processed images I1 are images obtained by applying different first image processing to an image of a latent fingerprint. The first image processing may include one or more image enhancement processing, latent processing, and feature extraction processing. The combination of one or more image enhancement processing, latent processing, and feature extraction processing differs for each first image processing. In other words, each of the multiple first processed images I1 has been subjected to a different combination of image processing.

[0045] For example, as described above, if there are L (=l1×l2×l3×l4) types of first image processing, the acquisition unit 211 can acquire a maximum of L types of first processed images I1.

[0046] The multiple second-processed images I2 are images obtained by applying different second image processing to an image of an inked fingerprint. The second image processing may include one or more image enhancement processing and feature extraction processing. In other words, the first image processing includes more image processing than the second image processing. The combination of one or more image enhancement processing and feature extraction processing differs for each second image processing. In other words, each of the multiple second-processed images I2 has a different combination of image processing applied to it.

[0047] For example, as described above, if there are T (= t1 × t2) types of second image processing, the acquisition unit 211 can acquire up to T types of second-processed images I2. The number of types of processing included in each of the image processing included in the first image processing is greater than the number of types of processing included in each of the image processing included in the second image processing. Therefore, the number of t1 types is smaller than the number of l1 types, and the number of t2 types is smaller than the number of l4 types. Furthermore, the number of T types is smaller than the number of L types.

[0048] An example of an index representing a matching characteristic (a property related to the matching result) corresponding to a matching pair is a matching score corresponding to the matching pair. In this embodiment, the matching score corresponding to the matching pair is used as an index representing a matching characteristic (a property related to the matching result) corresponding to the matching pair. In this embodiment, it may be said that the matching characteristic is represented by the matching score.

[0049] The matching unit 213 matches any one of the plurality of first-processed images with any one of the plurality of second-processed images and outputs each of a plurality of matching scores (step S21). If the acquisition unit 211 acquires L types of first-processed images I1 and T types of second-processed images I2, there are L×T (=N) types of matching pairs. In other words, the matching unit 213 may perform N types of matching.

[0050] The matching scores corresponding to each of the multiple (N types) matching pairs may be correlated with one another. The matching score correlation calculation unit 2141 calculates a matching score correlation indicating the correlation between the matching scores corresponding to each of the multiple matching pairs (step S22). For example, when matching I latent fingerprint images with J inked fingerprint images, I×J types of matching scores may be calculated for each matching pair. The matching score correlation calculation unit 2141 may calculate a matching score correlation indicating the correlation between the I×J types of matching scores. The matching score correlation calculation unit 2141 may express the matching score correlation as a numerical value between 0 and 1, for example. The matching score correlation calculation unit 2141 may calculate a matching score correlation indicating the correlation between each of the multiple matching scores that are equal to or greater than a predetermined value. In other words, the matching score correlation calculation unit 2141 may exclude matching scores that are less than the predetermined value from the calculation of the matching score correlation.

[0051] The distance matrix calculation unit 2142 calculates the distance between the matching characteristics based on the matching score correlation (step S23). The distance matrix calculation unit 2142 calculates the distance so that the closer the matching characteristics are to each other (the larger the value of the matching score correlation), the closer the distance is. The distance matrix calculation unit 2142 calculates the distance so that the less the matching characteristics are correlated (the smaller the value of the matching score correlation), the farther the distance is. The distance matrix calculation unit 2142 may calculate the distance, for example, as the difference between a constant 1 and the numerical value of the matching score correlation expressed as a value between 0 and 1. The distance matrix calculation unit 2142 may calculate a distance matrix of, for example, (N-1) x (N-1).

[0052] The conversion unit 215 converts the distance matrix into two-dimensional information (step S24). The two-dimensional information indicates the positional relationship of each matching characteristic in two-dimensional space. In other words, the conversion unit 215 reduces the number of dimensions of the distance matrix. The conversion unit 215 may perform dimension reduction using, for example, multi-dimensional scaling (MDS). Alternatively, the conversion unit 215 may perform dimension reduction using UMAP (Uniform Manifold Approximation and Projection).

[0053] For example, FIG. 4A illustrates the positional relationship of each matching characteristic in two-dimensional space. The distance between the matching characteristics in two-dimensional space allows understanding of the correlation between the matching characteristics. Matching characteristics that are close to each other in two-dimensional space are strongly correlated. Matching characteristics that are far from each other in two-dimensional space are weakly correlated.

[0054] The identification unit 216 identifies a plurality of search-target match pairs based on the match score correlation (step S25). The identification unit 216 identifies a plurality of search-target match pairs based on the positional relationship of each of the match characteristics in the two-dimensional space indicated by the two-dimensional information. The identification unit 216 may eliminate match pairs corresponding to match characteristics that are located relatively close to each other in the two-dimensional space. In other words, the identification unit 216 reduces the number of match pairs by eliminating the match pairs.

[0055] The output unit 212 outputs a plurality of search target matching pairs to be used in the search process from among the matching pairs that match any one of the plurality of first processed images with any one of the plurality of second processed images (step S26).

[0056] The output unit 212 outputs the type of image processing (each of enhancement and feature extraction may be referred to as image processing) included in the first image processing.

[0057] The output unit 212 may output a two-dimensional image showing the positional relationship of the matching characteristics in two-dimensional space and a two-dimensional image showing the positional relationship of the matching characteristics corresponding to the search target matching pair after the match pairs have been reduced. In other words, the output unit 212 may output a visualization of the distribution of the matching characteristics before and after processing. [2-5-1: Example 1 of Operation for Identifying Search Target Matching Pairs]

[0058] In step S25, the identification unit 216 may divide a two-dimensional image showing the positional relationship of the matching characteristics in two-dimensional space into a plurality of regions, and identify a plurality of search-target match pairs from each of the plurality of regions. For example, as illustrated in FIG. 4( b), the identification unit 216 may divide a two-dimensional image showing the positional relationship of the matching characteristics in two-dimensional space into a plurality of grid-shaped regions. The identification unit 216 may identify at least one search-target match pair from the divided plurality of grid-shaped regions. [2-5-1: Example 2 of Operation for Identifying Search-target Match Pairs]

[0059] In step S25, when the matching characteristics are distributed at a density equal to or higher than a predetermined density to form a group in a two-dimensional image showing the positional relationship of the matching characteristics in two-dimensional space, the identification unit 216 may thin out matching pairs corresponding to at least some of the matching characteristics in the group. For example, the identification unit 216 may thin out matching pairs corresponding to matching characteristics that are distributed at a relatively high density among the matching characteristics located in the two-dimensional space. For example, the identification unit 216 may thin out matching pairs corresponding to matching characteristics distributed in area C illustrated in FIG. 4( c). Conversely, the identification unit 216 may identify, as search-target matching pairs, matching pairs corresponding to matching characteristics that are distributed at a relatively low density among the matching characteristics located in the two-dimensional space. The identification unit 216 may identify multiple search-target matching pairs by eliminating matching pairs with similar matching characteristics. [2-6: Technical Effects of the Information Processing Device 2]

[0060] The information processing device 2 disclosed herein identifies multiple search target match pairs based on match score correlation, thereby achieving high reliability of the identifying operation. The information processing device 2 reduces the dimension of information based on match score correlation, thereby realizing a reduction in match pairs without limiting the search range. Furthermore, by reducing the dimension of information based on match score correlation and converting it into visualized information before identifying multiple search target match pairs, the user can easily understand the difference between match pairs before reduction and match pairs after reduction. Furthermore, the information processing device 2 identifies multiple search target match pairs from pairs of people who can be matched with a predetermined or higher matching accuracy, thereby preventing search processing from being performed on match pairs that are not suitable for matching in the first place. [3: Third Embodiment]

[0061] A third embodiment of an information processing device, an information processing method, and a recording medium will be described. Hereinafter, a third embodiment of an information processing device, an information processing method, and a recording medium will be described using an information processing device 3 according to this disclosure.

[0062] In the third embodiment, the acquisition operation by the acquisition unit 311 (i.e., the operation of step S20 shown in FIG. 3) is different. [3-1: Information Processing Operation Performed by Information Processing Device 3]

[0063] As shown in FIG. 6 , the acquisition unit 311 selects and acquires a plurality of first processed images from among a plurality of processed images obtained by applying different first image processes to a first pattern image, based on features detected from each of the plurality of processed images (step S201). The detected features include information about feature points detected from the processed images. The acquisition unit 311 may select and acquire a plurality of pieces of first image processed information based on at least one of whether the positions of feature points included in each of the plurality of processed images match or do not match. The acquisition unit 311 may select a first processed image to acquire based on the number of feature points included in the processed image that do not match the positions of feature points included in other processed images. In other words, the acquisition unit 311 selects and acquires a plurality of pieces of first image processed information based on the degree of overlap between the positions of feature points included in each of the plurality of processed images. In other words, the acquisition unit 311 selects and acquires a plurality of pieces of first image processed information based on the appearance tendency of feature points appearing in each of the plurality of processed images. That is, the acquisition unit 311 acquires a plurality of first processed images from the plurality of processed images, excluding processed images for which similar processed images exist.

[0064] Furthermore, the acquisition unit 311 selects and acquires a plurality of second processed images from a plurality of processed images obtained by applying different second image processes to the second pattern image, based on features detected from each of the plurality of processed images. The acquisition unit 311 may select and acquire a plurality of pieces of second image processed information based on at least one of a match and a mismatch in the positions of feature points contained in each of the plurality of processed images. In other words, the acquisition unit 311 may select and acquire a plurality of pieces of second image processed information based on the degree of overlap of the positions of feature points contained in each of the plurality of processed images. In other words, the acquisition unit 311 selects and acquires a plurality of pieces of second image processed information based on the appearance tendency of feature points contained in each of the plurality of processed images. In other words, the acquisition unit 311 acquires a plurality of second processed images from the plurality of processed images, excluding processed images for which similar processed images exist.

[0065] In this disclosure, the matching operation, calculation operation, and identification operation described in the second embodiment may be omitted. Even when only the selection operation described in the third embodiment is performed, the number of processed images to be acquired can be reduced, so that the number of matching pairs is reduced, and multiple search target matching pairs can be identified so that the corresponding matching results are not similar to each other. [3-2: Technical Effects of Information Processing Device 3]

[0066] The information processing device 3 according to this disclosure thins out similar processed images, and is therefore able to identify multiple search target match pairs with dissimilar match characteristics. As a result, it is also able to identify multiple search target match pairs whose respective match results are dissimilar. The information processing device 3 is able to realize appropriate preparation for search processing. Because the information processing device 3 thins out processed images, it is possible to reduce the number of match pairs and suppress the processing load of the match operation. [4: Fourth Embodiment]

[0067] Fourth Embodiment of Information Processing Apparatus, Information Processing Method, and Recording Medium Will Be Described Hereinafter, a fourth embodiment of an information processing apparatus, an information processing method, and a recording medium will be described using an information processing apparatus 4 according to this disclosure.

[0068] 7, the fourth embodiment differs from the second and third embodiments in that a search unit 417 is further implemented in the calculation device 21. The fourth embodiment also differs from the second and third embodiments in the operation of the output unit 412. [4-1: Information Processing Operation Performed by Information Processing Device 4]

[0069] The search unit 417 performs a search process using a genetic algorithm. The search unit 417 searches a plurality of search target matching pairs for a combination of matching pairs that can match the first pattern image with the second pattern image with a desired accuracy, based on the matching characteristics of each of the plurality of search target matching pairs. The search unit 417 may search for a matching pair combination that is optimal for matching the first pattern image with the second pattern image. In other words, the search unit 417 searches for a combination of image processing pairs that is optimal for matching the first pattern image with the second pattern image.

[0070] The output unit 412 outputs a combination of multiple matching pairs that are the search results. The output unit 412 outputs the multiple types of image processing included in the first image processing corresponding to the searched matching pair, and the multiple types of image processing included in the second image processing. A table may be implemented in the storage device 22 that stores matching pairs and the multiple types of image processing included in the first image processing corresponding to the matching pair, and the multiple types of image processing included in the second image processing, in association with each other. In this case, the output unit 412 may output the types of image processing by referring to this table. [4-2: Technical Effects of Information Processing Device 4]

[0071] The information processing device 4 according to this disclosure performs the search process after identifying a matching pair useful for the search process, thereby reducing the processing load. The information processing device 4 performs the search process using a genetic algorithm, thereby being able to search for the optimal combination of matching pairs. The information processing device 4 can contribute to the matching of latent fingerprints by searching for a combination of matching pairs that can achieve a high effect of fusion matching. The information processing device 4 outputs the type of image processing included in the first image processing and the type of image processing included in the second image processing, thereby allowing the user to understand which specific image processing has been applied to the processed image that is advantageous for fusion matching. [5: Supplementary Note]

[0072] The following supplementary notes are further disclosed in relation to the above-described embodiment. [Supplementary Note 1] An information processing device comprising: acquisition means for acquiring a plurality of first processed images obtained by applying a different first image processing to a first pattern image and a plurality of second processed images obtained by applying a different second image processing to a second pattern image; and output means for outputting a plurality of search-target match pairs to be used in a search process among match pairs for matching any one of the plurality of first processed images with any one of the plurality of second processed images, wherein the search process is a process for searching for a combination of the match pairs, and the plurality of search-target match pairs are identified so that the match characteristics of each of the plurality of search-target match pairs are not similar. [Supplementary Note 2] The information processing device according to Supplementary Note 1, comprising: matching means for outputting each of the match scores corresponding to each of the match pairs; calculation means for calculating a match score correlation indicating the correlation between each of the match scores; and identification means for identifying the plurality of search-target match pairs based on the match score correlation. [Supplementary Note 3] The information processing device according to Supplementary Note 1, wherein the second pattern image contains more features that can be used for matching pattern images compared to the first pattern image, and the first pattern image has more types of image processing performed when extracting features that can be used for matching pattern images compared to the second pattern image. [Supplementary Note 4] The information processing device according to Supplementary Note 2, wherein the first pattern image is an image of a latent fingerprint, the first image processing includes one or more image enhancement processing, latent processing, and feature extraction processing, and a combination of the one or more image enhancement processing, latent processing, and feature extraction processing differs for each first image processing, and the second pattern image is an image of an inked fingerprint, and the second image processing includes one or more image enhancement processing and feature extraction processing, and a combination of the one or more image enhancement processing and feature extraction processing differs for each second image processing. [Supplementary Note 5] The information processing device according to Supplementary Note 2, wherein the calculation means calculates a distance matrix based on the matching score correlation, and the information processing device includes a conversion means that converts the distance matrix into two-dimensional information indicating a positional relationship of the matching scores in a two-dimensional space, and the identification means identifies the plurality of search target match pairs based on the positional relationship of the matching scores in the two-dimensional space indicated by the two-dimensional information.[Supplementary Note 6] The information processing device according to Supplementary Note 2, wherein the acquisition means selects and acquires the plurality of first image processed information from among a plurality of processed images obtained by applying different first image processing to the first pattern image, based on features detected from each of the plurality of processed images. [Supplementary Note 7] The information processing device according to Supplementary Note 2, wherein the acquisition means selects and acquires the plurality of second image processed information from among a plurality of processed images obtained by applying different second image processing to the second pattern image, based on features detected from each of the plurality of processed images. [Supplementary Note 8] The information processing device according to Supplementary Note 5, wherein the identification means divides a two-dimensional image showing the positional relationship of the matching scores in the two-dimensional space into a plurality of regions, identifies at least one matching score from each of the plurality of regions, and identifies the plurality of search target matching pairs corresponding to the identified matching scores. [Supplementary Note 9] The information processing device according to Supplementary Note 8, wherein the identification means divides a two-dimensional image showing the positional relationship of the matching scores in the two-dimensional space into a plurality of grid-shaped regions. [Supplementary Note 10] The information processing device according to Supplementary Note 5, wherein the identification means, when the matching scores are distributed at a density equal to or higher than a predetermined value to form a group in a two-dimensional image showing the positional relationships of the matching scores in the two-dimensional space, identifies the plurality of search target match pairs by thinning out the match pairs corresponding to at least some of the matching scores among the match scores belonging to the group. [Supplementary Note 11] The information processing device according to Supplementary Note 6, wherein the detected features include information on feature points detected from the processed image, and the acquisition means selects and acquires the plurality of first image processed information according to at least one of a match and a mismatch in positions of feature points included in each of the plurality of processed images. [Supplementary Note 12] The information processing device according to Supplementary Note 7, wherein the detected features include information on feature points detected from the processed image, and the acquisition means selects and acquires the plurality of second image processed information according to at least one of a match and a mismatch in positions of feature points included in each of the plurality of processed images.[Supplementary Note 13] The information processing device according to Supplementary Note 2, further comprising a search means for performing the search process, wherein the search means searches the plurality of search-target match pairs for a combination of match pairs that allows the first pattern image and the second pattern image to be matched with a desired accuracy based on the matching characteristics of each of the plurality of search-target match pairs. [Supplementary Note 14] The information processing device according to Supplementary Note 13, wherein the search means performs the search process using a genetic algorithm. [Supplementary Note 15] The information processing device according to Supplementary Note 13, wherein the output means outputs a type of image processing included in the first image processing and a type of image processing included in the second image processing that correspond to the match pair included in the combination searched for by the search means. [Supplementary Note 16] The information processing device according to Supplementary Note 5, wherein the output means outputs a two-dimensional image indicating the positional relationship of the match scores in the two-dimensional space. [Supplementary Note 17] The information processing device according to Supplementary Note 2, wherein the calculation means calculates the match score correlation corresponding to each of the match pairs whose match scores are equal to or greater than a predetermined value. [Supplementary Note 18] An information processing method comprising: obtaining a plurality of first processed images obtained by performing a first image processing different from each other on a first pattern image; and obtaining a plurality of second processed images obtained by performing a second image processing different from each other on a second pattern image; outputting a plurality of search-target matching pairs to be used in a search process from among matching pairs that match any one of the plurality of first processed images with any one of the plurality of second processed images; the search process is a process for searching for a combination of the matching pairs; and specifying the plurality of search-target matching pairs so that the matching characteristics of each of the plurality of search-target matching pairs are not similar.[Supplementary Note 19] A recording medium having recorded thereon a computer program for causing a computer to execute an information processing method, the method comprising: acquiring a plurality of first processed images obtained by performing a first image processing different from each other on a first pattern image, and a plurality of second processed images obtained by performing a second image processing different from each other on a second pattern image; outputting a plurality of search-target matching pairs to be used in a search process from among matching pairs that match any one of the plurality of first processed images with any one of the plurality of second processed images; the search process being a process for searching for a combination of the matching pairs; and specifying the plurality of search-target matching pairs so that the matching characteristics of each of the plurality of search-target matching pairs are not similar.

[0073] Although this disclosure has been described above with reference to the embodiments, this disclosure is not limited to the above-described embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of this disclosure within the scope of this disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate.

[0074] 1, 2, 3, 4 Information processing device 11, 211, 311 Acquisition unit 12, 212, 412 Output unit 213 Matching unit 214 Calculation unit 2141 Matching score correlation calculation unit 2142 Distance matrix calculation unit 215 Conversion unit 216 Identification unit 417 Search unit

Claims

1. An acquisition means for acquiring a plurality of first processed images obtained by performing different first image processes on a first pattern image respectively, and a plurality of second processed images obtained by performing different second image processes on a second pattern image respectively; and an output means for outputting a plurality of search target collation pairs used in a search process among collation pairs obtained by collating any one of the plurality of first processed images with any one of the plurality of second processed images, wherein the search process is a process of searching for combinations of the collation pairs, and the plurality of search target collation pairs are specified such that the collation characteristics of each of the plurality of search target collation pairs are not similar. An information processing apparatus.

2. A collation means for outputting each of the collation scores corresponding to each of the collation pairs; a calculation means for calculating a collation score correlation indicating the correlation of each of the collation scores; and a specifying means for specifying the plurality of search target collation pairs based on the collation score correlation. The information processing apparatus according to claim 1.

3. The second pattern image contains more features that can be used for collation between pattern images as compared with the first pattern image, and the first pattern image has a greater variety of types of image processes performed when extracting features that can be used for collation between pattern images as compared with the second pattern image. The information processing apparatus according to claim 1.

4. The first pattern image is an image of a latent fingerprint, the first image process includes one or more image enhancement processes, latent processes, and feature extraction processes, and the combination of the one or more image enhancement processes, latent processes, and feature extraction processes is different for each first image process. The second pattern image is an image of a pressed fingerprint, the second image process includes one or more image enhancement processes and feature extraction processes, and the combination of the one or more image enhancement processes and feature extraction processes is different for each second image process. The information processing apparatus according to claim 2.

5. The calculation means calculates a distance matrix based on the collation score correlation, and includes a conversion means for converting the distance matrix into two-dimensional information indicating the positional relationship of the collation scores in a two-dimensional space. The specifying means specifies the plurality of search target collation pairs based on the positional relationship of the collation scores in the two-dimensional space indicated by the two-dimensional information. The information processing apparatus according to claim 2.

6. The acquisition means selects and acquires the plurality of first image - processed information based on features detected from each of the plurality of processed images obtained by performing different first image processes on the first pattern image, for the information processing apparatus according to claim 2.

7. The acquisition means selects and acquires the plurality of second image - processed information based on features detected from each of the plurality of processed images obtained by performing different second image processes on the second pattern image, for the information processing apparatus according to claim 2.

8. The specifying means divides a two - dimensional image showing the positional relationship of the collation scores in the two - dimensional space into a plurality of regions, specifies at least one of the collation scores from each of the plurality of regions, and specifies the plurality of search - target collation pairs corresponding to the specified collation scores, for the information processing apparatus according to claim 5.

9. The specifying means divides a two - dimensional image showing the positional relationship of the collation scores in the two - dimensional space into a plurality of grid - like regions, for the information processing apparatus according to claim 8.

10. In the two - dimensional image showing the positional relationship of the collation scores in the two - dimensional space, when the collation scores are distributed with a density equal to or higher than a predetermined value and form a group, the specifying means thins out the collation pairs corresponding to at least some of the collation scores belonging to the group to specify the plurality of search - target collation pairs, for the information processing apparatus according to claim 5.

11. The detected feature includes information regarding feature points detected from the processed image, and the acquisition means selects and acquires the plurality of first image - processed information according to at least one of the coincidence and non - coincidence of the positions of the feature points included in each of the plurality of processed images, for the information processing apparatus according to claim 6.

12. The detected feature includes information regarding feature points detected from the processed image, and the acquisition means selects and acquires the plurality of second image - processed information according to at least one of the coincidence and non - coincidence of the positions of the feature points included in each of the plurality of processed images, for the information processing apparatus according to claim 7.

13. The information processing apparatus according to claim 2, further comprising search means for performing the search process, wherein the search means searches for a combination of the collation pairs that enables collation of the first pattern image and the second pattern image with a desired accuracy based on the collation characteristics of each of the plurality of search target collation pairs from the plurality of search target collation pairs.

14. The information processing apparatus according to claim 13, wherein the search means performs the search process using a genetic algorithm.

15. The information processing apparatus according to claim 13, wherein the output means outputs the types of image processing included in the first image processing corresponding to the collation pairs included in the combination searched by the search means and the types of image processing included in the second image processing.

16. The information processing apparatus according to claim 5, wherein the output means outputs a two-dimensional image indicating the positional relationship of the collation scores in the two-dimensional space.

17. The information processing apparatus according to claim 2, wherein the calculation means calculates the collation score correlation corresponding to each of the collation pairs for which the collation score is equal to or greater than a predetermined value.

18. An information processing method, comprising: obtaining a plurality of first processed images obtained by performing different first image processes on a first pattern image, and a plurality of second processed images obtained by performing different second image processes on a second pattern image; outputting a plurality of search target collation pairs used in a search process from among collation pairs that collate any one of the plurality of first processed images with any one of the plurality of second processed images, wherein the search process is a process of searching for a combination of the collation pairs, and the plurality of search target collation pairs are specified such that the collation characteristics of each of the plurality of search target collation pairs are not similar.

19. A recording medium having recorded thereon a computer program for causing a computer to execute an information processing method, the method comprising: obtaining a plurality of first processed images obtained by performing different first image processes on a first pattern image, and a plurality of second processed images obtained by performing different second image processes on a second pattern image; outputting a plurality of search target collation pairs used in a search process from among collation pairs that collate any one of the plurality of first processed images with any one of the plurality of second processed images, wherein the search process is a process of searching for a combination of the collation pairs, and the plurality of search target collation pairs are specified such that the collation characteristics of each of the plurality of search target collation pairs are not similar.

Images