JP6692309B2 - Biometric authentication device, biometric authentication method, and biometric authentication program - Google Patents

Description

  The present invention relates to a biometric authentication device, a biometric authentication method, and a biometric authentication program.

  2. Description of the Related Art In recent years, biometric authentication that authenticates a living body based on a biometric image obtained by capturing a part of the living body has become popular. For example, if the biometric image is an image of a human face, face landmarks such as eyes and nose are present, and thus the face shape in the biometric image can be normalized by using a flatting technique or the like. On the other hand, if the biometric image is a palm image, even if a flatting technique is used as in the case of a face image, the shape unique to the palm causes deformation that is inconsistent with the structure of the hand. There are cases. Therefore, a technique has been proposed that focuses on the structure of the base of the finger of the palm and normalizes the palm image (see, for example, Patent Document 1). In addition, a method of using triangulation for locating a feature point included in a biometric image has been proposed (see, for example, Patent Document 2).

JP, 2005-028724, A Japanese Patent Publication No. 2005-519625

  However, in the above-described conventional technique, the palm is not limited to the base of the finger, and since there are many moving parts, the reproducibility of the normalization of the palm shape is low, and therefore the palm feature points are stably extracted. Therefore, there is a problem that the accuracy of palm vein authentication is low.

  The disclosed technique is made in view of the above, and an object thereof is to provide, for example, a biometric authentication device, a biometric authentication method, and a biometric authentication program that improve the accuracy of palm vein authentication.

In an example of the disclosed technology, a biometric authentication device, an image acquisition unit that acquires an image of a biometric part used for biometric authentication, an extraction unit that extracts a processing target region from the acquired image acquired by the image acquisition unit , and Based on the estimation unit that estimates the three-dimensional shape of the living body part from the processing target region extracted by the extraction unit, and the three-dimensional shape of the living body part estimated by the processing target region and the estimation unit, While maintaining the distance between the vertices of each mesh data in the three-dimensional mesh data, the mesh generation unit that generates the three-dimensional mesh data of the living body part from the acquired image and the three-dimensional mesh data generated by the mesh generation unit. Mesh flattening processing unit for converting into two-dimensional mesh data, and the three-dimensional mesh data before conversion by the mesh flattening processing unit Curved surface flattening processing for mapping the brightness value of each point of the living body portion on the acquired image to each point on the two-dimensional mesh data by geometric transformation determined from the correspondence relationship with the two-dimensional mesh data after transformation. And a section.

  According to an example of the disclosed technology, the accuracy of palm vein authentication can be improved.

FIG. 1 is a diagram illustrating an example of a biometric authentication device according to an embodiment. FIG. 2 is a diagram illustrating an example of an image acquired by the biometric authentication device according to the embodiment. FIG. 3 is a diagram illustrating an example of the ROI extraction result by the biometric authentication device according to the embodiment. FIG. 4 is a diagram illustrating an example of a grayscale image for estimating the three-dimensional shape of the palm by the biometric authentication device according to the embodiment. FIG. 5 is a diagram showing an example in which the contour of the palm ROI and the contour of the three-dimensional shape estimation image are represented by a point group and a vertex set by the biometric authentication device according to the embodiment. FIG. 6 is a diagram illustrating an example in which a set of vertices is plotted in a three-dimensional space by the biometric authentication device according to the embodiment. FIG. 7 is a diagram illustrating an example of a three-dimensional mesh formed of Faces by triangulation of a vertex set by the biometric authentication device according to the embodiment. FIG. 8 is a diagram illustrating an outline of an example of a mesh flattening process including conversion from a three-dimensional mesh to a converted two-dimensional mesh by the biometric authentication device according to the embodiment. FIG. 9 is a diagram showing an example of a three-dimensional mesh before conversion, which is a two-dimensional projection of the three-dimensional mesh according to the embodiment. FIG. 10 is a diagram illustrating an example of a converted two-dimensional mesh obtained by converting the three-dimensional mesh into a two-dimensional shape by the biometric authentication device according to the embodiment. FIG. 11 is a diagram illustrating an example in which the pre-conversion three-dimensional mesh of the palm image according to the embodiment is superimposed on the acquired image. FIG. 12 is a diagram showing an example in which the converted two-dimensional mesh of the palm image converted by the geometric conversion by the biometric authentication device according to the embodiment is superimposed on the acquired image. FIG. 13 is a flowchart illustrating an example of image processing in the biometric authentication device according to the embodiment.

  Hereinafter, embodiments of a biometric authentication device, a biometric authentication method, and a biometric authentication program disclosed in the present application will be described in detail with reference to the drawings. The biometric authentication device, biometric authentication method, and biometric authentication program disclosed in the present application are not limited by the following embodiments. Further, the following embodiments and modified examples can be appropriately combined without departing from the spirit of the invention without departing from the contradiction. In addition, the same reference numerals are given to the same elements in the respective drawings to be illustrated and described in the drawings, and the description of the later-described elements will be appropriately omitted.

[Biometric Authentication Device According to Embodiment]
FIG. 1 is a diagram illustrating an example of a biometric authentication device according to an embodiment. The biometric authentication device 1 according to the embodiment is a so-called palm vein authentication device, and includes an image processing device 100, an imaging element 200, a light emitting element 300, and an authentication processing device 400. The light emitting element 300 is, for example, an LED (Light Emitting Diode) that irradiates the palm of the user with near infrared light. The image pickup device 200 acquires a palm image of the user based on the near-infrared light emitted from the light emitting device 300 and reflected by the palm of the user. For example, a CMOS (Complementary Metal Oxide Semiconductor) For example, an image sensor.

  The image processing apparatus 100 generates an image for extracting a feature point of a palm vein from the palm image acquired by the image sensor 200. Details of the image processing apparatus 100 will be described later. The authentication processing device 400 generates a palm vein pattern of the user from the image generated by the image processing device 100, and authenticates the user based on the generated palm vein pattern of the user. Illustration and description of the configurations of the biometric authentication device 1 other than the image processing device 100, the imaging element 200, the light emitting element 300, and the authentication processing device 400 are omitted.

[Image Processing Device According to Embodiment]
The image processing device 100 is a processing device such as an MPU (Micro Processing Unit) and includes an input unit 110 and a curved surface flattening unit 120. The input unit 110 includes an image acquisition unit 111, a ROI (Region Of Interest) extraction processing unit 112, and a three-dimensional shape estimation unit 113.

  The image acquisition unit 111 controls the image sensor 200 and the light emitting element 300 to acquire an image with the palm of the user as a subject. The image acquisition unit 111 and the image sensor 200 are an example of an image acquisition unit. FIG. 2 is a diagram illustrating an example of an image acquired by the biometric authentication device according to the embodiment. FIG. 2 shows a VGA size (640 × 480 pixels) grayscale image as an example of the acquired image 51.

  The ROI extraction processing unit 112 generates an ROI extraction image 52 in which the palm ROI 52a is extracted from the acquired image 51 acquired by the image acquisition unit 111. FIG. 3 is a diagram illustrating an example of the ROI extraction result by the biometric authentication device according to the embodiment. As shown in FIG. 3, the ROI extraction image 52 includes a palm ROI 52a as the palm region of interest included in the acquired image 51. The ROI (Region Of Interest) refers to a region of interest, and in the embodiment, is a palm region to be processed in the palm image of the acquired image 51. The ROI extraction processing unit 112 extracts the palm ROI 52a using an existing method such as image segmentation.

  The three-dimensional shape estimation unit 113 estimates the three-dimensional shape of the subject in the palm ROI 52a of the acquired image 51. The three-dimensional shape estimation unit 113 estimates the three-dimensional shape of the subject in the ROI extraction image 52 from the grayscale image by using a known method such as SFS (Shape-From Shade) method. FIG. 4 is a diagram illustrating an example of a grayscale image for estimating the three-dimensional shape of the palm by the biometric authentication device according to the embodiment. In the gray scale image 53 shown in FIG. 4, in the three-dimensional shape estimation image 53a corresponding to the palm ROI of the grace case, the gradation is mapped to each pixel by the calibration of the image sensor 200 or the like with respect to the palm of the subject. By doing so, the side closer to the image sensor 200 is represented by a gradation showing lower brightness, and the side farther from the image sensor 200 is represented by a gradation showing higher brightness.

  The three-dimensional shape may be externally acquired as an input value via an API (Application Program Interface). Such input values can be obtained from, for example, Kinect for Windows (registered trademark), stereo camera, dual camera, random dot method, three-dimensional spectrum light, laser autofocus, or the like. In this case, the input unit 110 has a “three-dimensional shape input unit” instead of the three-dimensional shape estimation unit 113.

  The curved surface flattening unit 120 includes a mesh generation unit 121, a mesh flattening processing unit 122, and a curved surface flattening processing unit 123.

  The Mesh generation unit 121 is an example of a discretization processing unit that divides the input three-dimensional palm image data into meshes and discretizes them. The Mesh generation unit 121 generates a mesh (mesh (mesh data)) by dividing the area of the three-dimensional data. In the embodiment, by including the element of the palm, which is a point in the process of deforming the palm, such as the base of the thumb of the palm of the human hand, in the mesh as the vertex element, the subsequent shape normalization can be performed accurately.

  Although there are various formats for the mesh, a triangular mesh is used in the embodiment. Triangular Mesh is composed of Vertex, a set of triangle vertices, and Face, which is a set of three sides of a triangle.

  The Mesh generation unit 121 generates a Mesh from the palm ROI 52a and each contour of the three-dimensional shape estimation image 53a.

  First, the Mesh generation unit 121 expresses the outline of the palm ROI 52a by the point group G1. Then, the mesh generation unit 121 determines that the points where the change in the contour shape of the palm ROI 52a (for example, the change in the area, the change in the length of the contour line, etc.) is less than or equal to a predetermined threshold even if the mesh is excluded from the point group G1. The vertex set V1 excluded from is generated. The mesh generation unit 121 also represents the contour of the three-dimensional shape estimation image 53a with a point group G2. Then, the mesh generation unit 121 has a point that the change of the contour shape of the three-dimensional shape estimation image 53a (for example, the change of the area, the change of the length of the contour, etc.) is less than or equal to a predetermined threshold even if the mesh generation unit 121 excludes the point group G2. Is generated from the point group G2 to generate a vertex set V2.

  FIG. 5 is a diagram showing an example in which the contour of the palm ROI and the contour of the three-dimensional shape estimation image are represented by a point group and a vertex set by the biometric authentication device according to the embodiment. The point group G1 representing the contour of the palm ROI 52a is indicated by a solid line in the outer contour shown in FIG. The vertex set V1 representing the contour of the palm ROI 52a is indicated by a chain line in the outer contour shown in FIG. The vertex set V1 is a point group generated by thinning out some points from the point group G1. The point group G2 representing the contour of the three-dimensional shape estimation image 53a is indicated by a solid line in the inner contour shown in FIG. The vertex set V2 representing the contour of the three-dimensional shape estimation image 53a is indicated by a chain line in the inner contour shown in FIG. The vertex set V2 is a point group generated by thinning out some points from the point group G2.

  Next, the Mesh generation unit 121 takes the vertex set V which is the union of the vertex sets V1 and V2. However, in the vertex set V, one of the two vertices whose mutual distances are equal to or less than a predetermined value may be excluded from the vertex set V.

  Next, the mesh generation unit 121 generates a three-dimensional mesh_M1 (see FIG. 7) including a face by the Delaunay triangulation method based on the vertex set V. Face generation by the Delaunay triangulation method is a well-known method described in Document 1, “M. de Berg et al.,“ Computational Geometry, 2nd ed. ”, Springer, 1998” and the like.

  Note that FIG. 6 is a diagram showing an example in which a vertex set is plotted in a three-dimensional space by the biometric authentication device according to the embodiment. In FIG. 6, the heights of the vertices included in the vertex set V are interpolated from the height information of adjacent points in the three-dimensional shape estimated by the three-dimensional shape estimation unit 113. In addition, FIG. 7 is a diagram illustrating an example of a three-dimensional mesh formed of Faces by triangulation of the vertex set by the biometric authentication device according to the embodiment.

  The mesh flattening processor 122 converts three-dimensional data into two-dimensional data. That is, the mesh flattening processing unit 122 uses the ISOMAP (Isometric Feature Mapping) method or the like to maintain the distance structure of the three-dimensional mesh (distance ratio between points, etc.) and convert the three-dimensional Mesh_M1 into the two-dimensional Mesh_M2 (hereinafter referred to as conversion). After that, it is converted to 2D Mesh_M2). FIG. 8 is a diagram illustrating an outline of an example of a mesh flattening process including conversion from a three-dimensional mesh to a converted two-dimensional mesh by the biometric authentication device according to the embodiment.

  The conversion from the three-dimensional mesh to the two-dimensional mesh by the ISOMAP method is well known as described in Reference 2, “J. Tenenbaum et al.,“ A global geometric framework for nonlinear dimensionality reduction, ”Science, 290: 2319-2323, 2000”. It is a technique. The ISOMAP method is an example of “conversion that maintains the distance structure of the three-dimensional mesh”, but any conversion method that can “maintain the distance structure of the three-dimensional mesh” is not limited to the ISOMAP method. ..

  Here, the two-dimensional projection of the three-dimensional Mesh_M1 for comparison with the post-conversion two-dimensional Mesh_M2 is referred to as the pre-conversion three-dimensional Mesh_M3. The pre-conversion 3D Mesh_M3 is a projection of the 3D Mesh_M1 on the same plane as the post-conversion 2D Mesh_M2. FIG. 9 is a diagram showing an example of a three-dimensional mesh before conversion, which is a two-dimensional projection of the three-dimensional mesh according to the embodiment. FIG. 10 is a diagram illustrating an example of a converted two-dimensional mesh obtained by converting the three-dimensional mesh into a two-dimensional shape by the biometric authentication device according to the embodiment.

  Comparing the three-dimensional mesh_M3 before conversion shown in FIG. 9 with the two-dimensional mesh_M2 after conversion shown in FIG. 10, for example, conversion is performed for Faces corresponding to the movable part of the palm such as the base of the thumb or the base of each finger. The transformed 2D Mesh_M2 is larger than the 3D mesh_M3 before the image while maintaining the distance ratio between each point of each Fece, and the geometric distortion such as projective distortion and the distortion due to the optical system are generated. Has been reduced. That is, by using “conversion that keeps the distance structure of the three-dimensional mesh as much as possible” such as ISOMAP method, the conversion distortion at the time of conversion from the three-dimensional Mesh_M1 to the two-dimensional mesh is reduced and the three-dimensional mesh It is possible to perform highly accurate conversion into two dimensions while maintaining the characteristics of the shape.

  The curved surface flattening unit 123 flattens the curved surface of the three-dimensional palm into two dimensions. That is, the curved surface flattening processing unit 123 determines the geometric transformation T from the three-dimensional Mesh_M1 to the two-dimensional mesh_M2 after conversion from the correspondence relationship between the three-dimensional Mesh_M1 and the two-dimensional mesh_M2 after conversion. Then, the curved surface flattening processing unit 123 extends the geometric transformation T to an arbitrary point on the palm, and normalizes the luminance value of the arbitrary point on the palm continuously mapped to a two-dimensional luminance value. Generate an image. Finally, the curved surface flattening processing unit 123 corrects the shape of the normalized image of the generated palm image, and outputs it to the authentication processing device 400.

  For example, the curved surface flattening processing unit 123 defines Thin-Plate Spline (TPS) transformation as the geometric transformation T. The Thin-Plate Spline transformation as the geometric transformation T is performed by converting each Face of the three-dimensional Mesh_M1 before conversion and each Face of the two-dimensional Mesh_M2 after conversion from the correspondence relationship before and after conversion of each Face in the two-dimensional Mesh_M2 after conversion. This is a conversion in which each face of the two-dimensional Mesh_M2 is smoothly connected after the conversion. The curved surface flattening processing unit 123 can perform geometric conversion with high accuracy by thin-plate spline conversion.

  Alternatively, the curved surface flattening processing unit 123 may determine, as the geometric transformation T, the Affine transformation for each Face corresponding to the three-dimensional Mesh_M1 and the transformed two-dimensional Mesh_M2. The Affine transformation as the geometric transformation T is uniquely determined from the correspondence between the three vertices of each Face of the three-dimensional Mesh_M1 and the transformed two-dimensional Mesh_M2, and transforms all points of the Face. When performing the Affine transformation as the geometric transformation T, the curved surface flattening processing unit 123 can perform the computation at high speed and can easily use hardware assisted computation such as GPU (Graphics Processing Unit). Therefore, the Affine transformation as the geometric transformation T is suitable for mounting on an embedded device or the like.

  FIG. 11 is a diagram illustrating an example in which the pre-conversion three-dimensional mesh of the palm image according to the embodiment is superimposed on the acquired image. FIG. 12 is a diagram showing an example in which the converted two-dimensional mesh of the palm image converted by the geometric conversion by the biometric authentication device according to the embodiment is superimposed on the acquired image. The post-conversion two-dimensional Mesh_M2 shown in FIG. 12 is reduced in conversion distortion at the time of conversion from the three-dimensional Mesh_M1 to the two-dimensional Mesh as compared with the pre-conversion three-dimensional Mesh_M3 shown in FIG. It has been converted to a two-dimensional mesh while maintaining its characteristics. From this, the post-conversion two-dimensional Mesh_M2 is compared with the pre-conversion three-dimensional Mesh_M3 by mapping the luminance value of each point of the palm from the three-dimensional Mesh_M1 by the geometrical conversion T, so that the palm feature amount (eg, movable) can be changed. It is possible to generate a normalized image in which the feature amount of a part) can be extracted with high accuracy. Therefore, according to the embodiment, the reproducibility of the palm image can be improved, and the characteristic amount of the movable portion according to the shape peculiar to the palm can be extracted efficiently and accurately at high speed.

  The authentication processing device 400 receives the normalized image shown in FIG. 12 output from the curved surface flattening processing unit 123 as an input, extracts the feature amount of the palm from the input output image, and outputs the palm vein including the extracted feature amount. The user is authenticated by collating the pattern with a palm vein pattern for collation registered in advance in the authentication processing device 400 or an external device.

[Image Processing in Biometric Authentication Device According to Embodiment]
FIG. 13 is a flowchart illustrating an example of image processing in the biometric authentication device according to the embodiment. The biometric authentication device 1 executes image processing in the biometric authentication device every time the palm of a person is held over the image sensor 200 for photographing.

  First, the image acquisition unit 111 performs an image acquisition process (step S11). Next, the ROI extraction processing unit 112 performs ROI extraction processing (step S12). Next, the three-dimensional shape estimation unit 113 performs a three-dimensional shape estimation process (step S13).

  Next, the Mesh generation part 121 performs a Mesh generation process (step S14). Next, the mesh flattening processing unit 122 performs mesh flattening processing (step S15). Next, the curved surface flattening processing unit 123 performs curved surface flattening processing (step S16). When step S16 ends, the curved surface flattening processing unit 123 outputs the output image by the curved surface flattening processing to the authentication processing device 400.

  According to the above-described embodiment, in the palm vein authentication, the characteristic of the deformation of the palm due to the skeleton of the hand is efficiently absorbed during the normalization of the image, so that the accuracy of the two-dimensionalization of the three-dimensional palm image is improved. It is possible to improve the reproducibility and the stability of feature point extraction by increasing the processing speed, and improve the processing speed and the authentication accuracy.

[Modification of Embodiment]
(1) Generation of Face In the embodiment, the Mesh generation unit 121 represents the vertex set V1 representing the contour of the palm ROI 52a generated from the point group G1 representing the contour of the palm ROI 52a and the contour of the three-dimensional shape estimation image 53a. A face is generated based on the vertex set V that is a union with the vertex set V2 representing the contour of the three-dimensional shape estimation image 53a generated from the point group G2. However, the present invention is not limited to this, and the Mesh generation unit 121 may generate Face based on the vertex set V1 representing the contour of the palm ROI 52a generated from the point group G1 representing the contour of the palm ROI 52a. Alternatively, the Mesh generation unit 121 may generate Face based on the vertex set V2 representing the contour of the 3D shape estimation image 53a generated from the point group G2 representing the contour of the 3D shape estimation image 53a.

(2) Program In the embodiment, the image processing in the biometric authentication device 1 is performed by the image processing device 100 included in the biometric authentication device 1. However, the image processing device 100 is not limited to this, and is provided as an image processing program and a CPU (Central Processing). It may be installed in a computer device having a processing device such as a Unit) and executed by the processing device of the computer device. The computer device may be a local computer or a remote computer installed in a data center. For example, the image processing according to the embodiment may be provided as a cloud service by a remote computer.

(3) Biometric Information In the embodiment, the biometric information is a palm vein pattern, but the biometric information is not limited to this, and may be any physical characteristic of a person such as a face or a fingerprint that can be used for authentication.

  The integration and distribution of each element, each device, and each unit in the above embodiments are not limited to those illustrated and described above, and may be appropriately changed in design depending on the mounting efficiency or processing efficiency. Further, the processing order of each processing in the above embodiments may be appropriately changed or the processing may be omitted.

  The configurations of the units illustrated in the above embodiments can be changed or omitted to the extent that they do not depart from the technical scope of the portable device and the biometric authentication method according to the disclosed technique. In addition, the embodiments are merely examples, and the disclosed technology includes other embodiments in which various modifications and improvements are made based on the knowledge of those skilled in the art, including the embodiments described in the disclosure section.

1 Biometric authentication device 51 Acquired image 52 ROI extracted image 52a Palm ROI
53 gray scale image 53a three-dimensional shape estimation image 100 image processing device 110 input unit 111 image acquisition unit 112 ROI extraction processing unit 113 three-dimensional shape estimation unit 120 curved surface flattening unit 121 mesh generation unit 122 mesh flattening processing unit 123 curved flat surface Authentication processing unit 200 Imaging element 300 Light emitting element 400 Authentication processing device

Claims (6)

An image acquisition unit that acquires an image of a biometric part used for biometric authentication,
An extraction unit that extracts a processing target region from the acquired image acquired by the image acquisition unit ,
An estimation unit that estimates the three-dimensional shape of the living body part from the processing target region extracted by the extraction unit;
A mesh generation unit that generates three-dimensional mesh data of the living body portion from the acquired image based on the processing target region and the three-dimensional shape of the living body portion estimated by the estimation unit ;
A mesh flattening processing unit that converts the three-dimensional mesh data generated by the mesh generation unit into two-dimensional mesh data while maintaining the distance between the vertices of each mesh data in the three-dimensional mesh data;
The brightness value of each point of the living body part on the acquired image is determined by the geometric conversion determined by the correspondence between the three-dimensional mesh data before conversion and the two-dimensional mesh data after conversion by the mesh flattening processing unit, And a curved surface flattening processing unit that maps to each point on the two-dimensional mesh data. The biometric authentication device according to claim 1, wherein the geometric transformation is Thin-Plate Spline transformation. The biometric authentication device according to claim 1, wherein the geometric transformation is Affine transformation. The mesh generation unit includes a first point group forming a contour of a ROI (Region Of Interest) in the processing target region extracted from the acquired image, and a contour of the living body part indicated by the information of the three-dimensional shape. A point set of either one of the second point groups or the union set of the first point group and the second point group is used as a vertex set, and the points included in the vertex set are used as the basis. The biometric authentication device according to any one of claims 1 to 3, wherein the three-dimensional mesh data is generated. A biometric authentication method executed by a biometric authentication device, comprising:
The image acquisition unit of the biometric authentication device acquires an image of a biometric part used for biometric authentication,
The extraction unit of the biometric authentication device extracts a processing target area from the acquired image acquired by the image acquisition unit,
An estimation unit of the biometric authentication device estimates a three-dimensional shape of the biometric portion from the processing target region extracted by the extraction unit,
The mesh generation unit of the biometric authentication device generates three-dimensional mesh data of the biometric portion from the acquired image based on the processing target area and the three-dimensional shape of the biometric portion estimated by the estimation unit ,
A mesh flattening processing unit of the biometric authentication device converts the three-dimensional mesh data generated by the mesh generation unit into two-dimensional mesh data while maintaining a distance between vertices of each mesh data in the three-dimensional mesh data. Then
On the acquired image, the curved surface flattening processing unit of the biometric authentication device performs geometric transformation determined from the correspondence between the three-dimensional mesh data before the transformation by the mesh flattening processing unit and the two-dimensional mesh data after the transformation. The biometric authentication method, comprising: each process of mapping the luminance value of each point of the biometric part to each point on the two-dimensional mesh data. Computer,
An image acquisition unit that acquires an image of a biometric part used for biometric authentication,
An extraction unit that extracts a processing target region from the acquired image acquired by the image acquisition unit ,
An estimation unit that estimates the three-dimensional shape of the living body part from the processing target region extracted by the extraction unit;
A mesh generation unit that generates three-dimensional mesh data of the living body portion from the acquired image based on the processing target region and the three-dimensional shape of the living body portion estimated by the estimation unit ;
A mesh flattening processing unit that converts the three-dimensional mesh data generated by the mesh generation unit into two-dimensional mesh data while maintaining the distance between the vertices of each mesh data in the three-dimensional mesh data;
The brightness value of each point of the living body part on the acquired image is determined by the geometric conversion determined by the correspondence between the three-dimensional mesh data before conversion and the two-dimensional mesh data after conversion by the mesh flattening processing unit, A biometric authentication program for functioning as a curved surface flattening processing unit that maps to each point on the two-dimensional mesh data.