JP4735104B2 - Fingerprint verification device, fingerprint verification processing program, and image verification method - Google Patents

Description

  The present invention relates to a fingerprint collation apparatus, a fingerprint collation processing program, and a fingerprint collation method for collating fingerprint image data.

  For example, as a technical means for authenticating whether or not the person is the person, the fingerprint image read for the identity verification is collated with a pre-registered fingerprint image to determine the similarity. A fingerprint image collating device that authenticates the user as a person when the degree of similarity is higher than a predetermined similarity is considered.

  In such a fingerprint image collation apparatus, as one method for determining the similarity between the fingerprint image (collation fingerprint image) read this time and the registered fingerprint image (registered fingerprint image), the same applicant can There is an image data collation method that has been applied (for example, see Patent Document 1).

  In this image data collation method, first, a plurality of rectangular areas (reference masks) are defined and arranged at predetermined positions on the collation fingerprint image, and each area having the same shape as each of the plurality of rectangular areas is placed on the registered fingerprint image. Set. Then, while changing the position of each setting area on the registered fingerprint image, the correlation coefficient between the verification fingerprint image and each area on the registered fingerprint image is obtained to obtain the maximum correlation coefficient. The degree of similarity is determined based on a comparison between the relative position distribution between the rectangular areas on the registered fingerprint image and the relative position distribution between the plurality of rectangular areas defined on the collation fingerprint image.

A plurality of rectangular areas (reference masks) defined and arranged on the collation fingerprint image are first defined and arranged on the registered fingerprint image, and a plurality of rectangular areas on which the maximum correlation coefficient is obtained are arranged on the collation fingerprint image. You may ask for.
JP 2000-194862 A

  As described in the conventional image data matching method, the plurality of rectangular areas defined first for the matching fingerprint image or the registered fingerprint image are predetermined with respect to a predetermined position based on the center of gravity position of the fingerprint image area. Are defined and arranged in the rectangular direction. For this reason, a plurality of rectangular areas are always present for a fingerprint image in which the image of the fingerprint differs depending on how the user's finger is placed on the fingerprint reading surface, or the fingerprint image itself has a large individual difference. There is a problem that it is difficult to say that it is arranged at an optimal position to capture the feature.

  The present invention has been made in view of such a problem. When a plurality of reference masks are arranged on a fingerprint image, the fingerprint can always be in an appropriate arrangement form to capture the characteristics of the fingerprint image. An object is to provide a collation device, a fingerprint collation processing program, and an image collation method.

The fingerprint collation apparatus according to claim 1 is based on a plurality of rectangular areas arranged in one fingerprint image and a plurality of maximum correlation areas on the other fingerprint image having the maximum correlation with each area image. A fingerprint collation device for collating fingerprints, wherein a rectangular area arrangement unit that arranges a plurality of rectangular areas on the one fingerprint image, and a fingerprint in each of the plurality of rectangular areas arranged by the rectangular area arrangement unit The ridge line direction analyzing means for analyzing the ridge line direction and a plurality of rectangular areas arranged by the rectangular area arranging means with respect to the fingerprint ridge line direction in each rectangular area analyzed by the ridge line direction analyzing means It is characterized in that a rectangular area relocation means for relocating across the longitudinal direction of the region.

Fingerprint matching apparatus according to claim 2, in fingerprint matching apparatus according to claim 1, the rectangular region temporary arrangement means for temporarily arranged by a plurality of rectangular regions in a grid shape before SL on one fingerprint image, A temporary placement ridge line direction analyzing unit that analyzes a ridge line direction of a fingerprint in each of the plurality of rectangular regions temporarily arranged in a grid pattern by the rectangular region temporary placement unit, and the rectangular region placement unit includes: Among the plurality of rectangular areas temporarily arranged in a square pattern by the rectangular area temporary arrangement means, a predetermined number of rectangular areas are analyzed in the respective rectangular areas analyzed by the temporary arrangement ridge line direction analyzing means. A feature is that a rectangular area whose ridgeline direction is the X direction and a rectangular area whose Y direction is the same are searched and arranged almost in half .

The fingerprint collation device according to claim 3 is the fingerprint collation device according to claim 1 or 2 , further comprising pressure distribution detection for detecting a pressure distribution of a fingerprint pressing pressure corresponding to each part of the one fingerprint image. And the rectangular area arranging means arranges a plurality of rectangular areas excluding a portion having a large pressure strain based on a fingerprint pressing pressure distribution in each part of one fingerprint image detected by the pressure distribution detecting means. It is a feature.

The fingerprint collation device according to claim 4 is the fingerprint collation device according to any one of claims 1 to 3 , further comprising a prescan that prescans and acquires an image on a reading surface of the fingerprint image. An image acquisition unit, a noise image presence determination unit that determines whether or not a noise image exists in the prescan image of the fingerprint reading surface acquired by the prescan image acquisition unit, and a fingerprint reading surface by the noise image presence determination unit Noise image removing means for removing the noise image from the fingerprint image read on the fingerprint reading surface when it is determined that a noise image is present in the pre-scanned image, the one fingerprint image and the other The fingerprint image is a fingerprint image from which the noise image has been removed by the noise image removing means.

The claim 1 (claim 5) a plurality of rectangular regions in the fingerprint collation apparatus (fingerprint collation processing program) (fingerprint matching method) on one of the fingerprint image according according to (Claim 6) of the present invention is arranged, The direction of the ridge line of the fingerprint in each of the images in the arranged rectangular areas is analyzed. Then, since the plurality of rectangular areas are rearranged so that the longitudinal direction of the rectangular area crosses the ridge line direction of the fingerprint in each rectangular area analyzed in the ridge line direction for the plurality of arranged rectangular areas, more in the plurality of rectangular areas. The feature portion of the fingerprint image can be captured, and more accurate fingerprint collation processing can be performed.

According to the fingerprint collation apparatus according to claim 2 of the present invention, in a fingerprint matching apparatus according to claim 1, temporarily placed in a plurality of rectangular regions in grid form on the hand of the fingerprint image, the The direction of the ridge line of the fingerprint in each of the images in the plurality of rectangular areas temporarily arranged in a grid pattern is analyzed. Then, in the rectangular area arrangement means, the ridge line direction of the fingerprint in each of the analyzed rectangular areas for the predetermined number of rectangular areas among the plurality of rectangular areas temporarily arranged in the grid pattern is the X direction. The rectangular area that becomes and the rectangular area that becomes the Y direction are searched and arranged almost in half, so that a predetermined number of fingerprint ridge lines are flowing at a position where the fingerprint ridge line flows in a good balance between the vertical and horizontal directions on one fingerprint image. A rectangular area can be arranged, and more accurate fingerprint collation processing can be performed.

According to the fingerprint collation apparatus according to claim 3 of the present invention, in the fingerprint collation apparatus according to claim 1 or 2 , in the rectangular area arrangement means, one fingerprint image detected by the pressure distribution detection means Based on the fingerprint pressure distribution in each part, a plurality of rectangular areas are arranged except for a part where the pressure distortion is large. Therefore, more accurate fingerprint collation processing is performed based on the rectangular area arranged in an area where the fingerprint image reading distortion is small. be able to.

According to a fingerprint collation apparatus according to claim 4 of the present invention, in the fingerprint collation apparatus according to any one of claims 1 to 3 , a fingerprint image reading surface image is further pre-scanned. And determining whether or not a noise image exists in the pre-scan image on the fingerprint reading surface. If it is determined that a noise image exists, the fingerprint obtained by reading the noise image on the fingerprint reading surface One fingerprint image to be verified and the other fingerprint image to be verified are fingerprint images from which this noise image has been removed, so that it is more accurate without being affected by noise such as residual fingerprints. Highly reliable fingerprint collation processing can be performed.

  Therefore, according to the present invention, when a plurality of reference masks are arranged on a fingerprint image, it is always possible to obtain an appropriate arrangement form for capturing the characteristics of the fingerprint image, and more accurate fingerprint collation processing is performed. It is possible to provide a fingerprint collation device, a fingerprint collation processing program, and an image collation method that can be performed.

  Embodiments of the present invention will be described below with reference to the drawings.

(First embodiment)
FIG. 1 is a block diagram showing a configuration of an electronic circuit of a fingerprint collation apparatus according to an embodiment of the present invention.

  This fingerprint collation device is composed of a control unit (CPU) 11, a storage device 12, a RAM 13, an image reading device 14, a display unit 15, an input unit 16, a storage medium reading unit 18, and a transmission control unit 19. Are connected to each other.

  In response to an input signal from the input unit 12, the control unit (CPU) 11 is a control program stored in advance in the storage device 12, or a storage medium such as a floppy disk drive FDD from an external storage medium 18a such as a floppy disk FD. The fingerprint collation processing program read into the storage device 12 via the reading unit 18 or the storage device 22 of the external computer terminal (program server 21) from the storage device 12 via the communication network 20 to the storage device 12 via the transmission control unit 19. The read fingerprint collation processing program is activated, and the operation of each part of the circuit is controlled using the RAM 13 as a work memory.

  The storage device 12 is constituted by a ROM of a semiconductor memory, a hard disk device HDD, or the like. In addition to storing the above-described control program read by the control unit (CPU) 11 immediately after turning on the power of the verification device, the storage device 12 reads an image. The registered fingerprint image Go read by the device 14 is stored.

  The RAM 13 is a work memory used by the control unit (CPU) 11 when executing the aforementioned control program.

  The image reading device 14 is a device for acquiring a registered fingerprint image Go and a collation fingerprint image Gs, and is an image sensor such as an image scanner device or a CCD (charge coupled device).

  The display unit 15 is a display device such as a CRT or a liquid crystal display that displays a result of collation between the registered fingerprint image Go and the collation fingerprint image Gs.

  The input unit 16 is an input device such as a keyboard device for the user of the verification device to instruct the control unit (CPU) 11 to issue a fingerprint image acquisition instruction or start a fingerprint verification process described later.

  FIG. 2 is a diagram showing a data memory secured in the RAM 13 of the fingerprint collation apparatus.

  The RAM 13 includes a read image memory 13a, a registered fingerprint image memory 13b, a collation fingerprint image memory 13c, a reference mask size / position information memory 13d, a maximum correlation position information memory 13e, a collation similarity memory 13f, a collation determination result memory 13g, and Other work memory areas are prepared.

  In the read image memory 13a, fingerprint image data (see FIGS. 3 and 4) read by the image reading device 14 is temporarily stored as a multi-tone image.

  The registered fingerprint image memory 13b stores and holds fingerprint image data Go (see FIG. 4B) of the registered person read by the image reading device 14 to be stored and registered in the storage device 12.

  In the collation fingerprint image memory 13c, fingerprint image data Gs (see FIGS. 3A and 4A) to be collated read by the image reading device 14 is stored as a multi-tone image.

  In the reference mask size / position information memory 13d, rectangular reference mask regions (templates) t0, t1, t2,..., Tn arranged on one fingerprint image data (Gs or Go) to be verified in the fingerprint verification apparatus. The mask size, arrangement position, and arrangement direction information (in the case of this embodiment, six of t0,..., T5: see FIG. 3 or FIG. 4A) are set and stored.

  In this reference mask size / position information memory 13d, the initial default arrangement position and the rectangular arrangement direction of each reference mask area t0,..., T5 are, for example, as shown in FIG. Gs or Go) is set in a grid-like layout position consisting of two vertical columns and three horizontal columns based on the barycentric position of Gs or Go).

  The initial default rectangular layout direction on the read fingerprint image (Gs or Go) is set to the vertically long direction, but the ridge line direction of the fingerprint image is analyzed for each reference mask region t0,..., T5. Then, as shown in FIG. 3B, the individual reference mask areas t0,..., T5 are selectively vertically and horizontally converted and set so that the ridgeline of the fingerprint image crosses as much as possible.

  In the maximum correlation position information memory 13e, for each image of the reference mask areas t0,..., T5 arranged on one fingerprint image data (Gs or Go) to be verified, the other fingerprint image data (Go or Gs), the position information of each maximum correlation (position) region k0,..., K5 (see FIG. 4B) where the maximum correlation coefficient is calculated is retrieved and stored.

  In the matching similarity memory 13f, the mutual positional relationships XYt1,..., XYt5 between t1, t2,..., T5 with respect to the reference mask area t0 for the one fingerprint image data (Gs or Go) and the other , K5 to the maximum correlation (position) region k0 on the fingerprint image data (Go or Gs) of the image, the matching similarity R is calculated according to the difference between the mutual positional relationships XYk1,. And memorized.

  In the collation determination result memory 13g, whether or not the registered fingerprint image Go and the collation fingerprint image Gs are evaluated according to whether the collation similarity R is higher than a preset similarity threshold THb is determined. The result is stored.

  FIG. 3 shows an initial arrangement state of reference mask areas (templates) t0, t1,..., T5 on a collation fingerprint image Gs accompanying a fingerprint collation process of the fingerprint collation device and a post-conversion arrangement state according to fingerprint image ridge line analysis. FIG.

  FIG. 4 shows the arrangement state of reference mask regions (templates) t0, t1,..., T5 on the collated fingerprint image Gs and the maximum correlation position (region) on the registered fingerprint image Go accompanying the fingerprint collation processing of the fingerprint collation device. ) K0, k1,..., K5.

  In this fingerprint collation device, when collating fingerprint images (Gs and Go), when arranging a plurality of rectangular reference mask areas t0, t1,..., T5 on one fingerprint image (Gs or Go), for example, FIG. As shown in FIG. 3 (A), the six rectangular reference mask areas t0, t1,..., T5 are all arranged vertically with reference to a predetermined position serving as the center of gravity of the fingerprint image, and then each mask area t0. , T1,..., T5 are analyzed, and the vertical and horizontal conversion is selectively performed in a direction in which at least one ridge line crosses for each mask area, that is, in a direction in which the longitudinal direction of the rectangle is orthogonal to the ridge line direction. Arranged.

  Thereby, a more characteristic image area on the fingerprint image Gs (or Go) to be collated can be recognized as the reference mask areas t0, t1,..., T5, and the reference mask areas t0, t1,. By collating images based on them, more accurate fingerprint collation / authentication processing can be performed.

  Next, the operation of the fingerprint collation device according to the first embodiment having the above configuration will be described.

  FIG. 5 is a flowchart showing fingerprint collation processing by the fingerprint collation apparatus 10.

  FIG. 6 is a flowchart showing a reference mask arrangement process accompanying the fingerprint collation process by the fingerprint collation apparatus 10.

  In this fingerprint collation processing, when the fingerprint image of the person to be collated is read by the image reading device 14 and stored in the read image memory 13a in the RAM 13, the read image is compared with the collation fingerprint image data Gs to the collation fingerprint image memory 13c. (Step S1).

  Then, the process proceeds to the reference mask arrangement process (step S2) in FIG. 6. First, as shown in FIG. 3 (A), the reference fingerprint image data Gs stored in the verification fingerprint image memory 13c is compared with the reference fingerprint image data Gs. Based on the initial reference mask information stored and defined in the mask size / position information memory 13d, six vertical rectangles of horizontal width a [dot] and vertical width b [dot] (a <b) are arranged in two vertical rows and horizontal rows. The reference mask areas t0, t1,..., T5 in the form of three rows of cells are arranged with reference to the barycentric position of the collation fingerprint image Gs (step S2a).

  Then, a sum S of absolute values of primary differential values which are differences between adjacent pixels in the X direction and the Y direction is calculated for the individual image data in each of the reference mask regions t0, t1, to t5. (Step S2b) In each of the reference mask areas t0, t1,..., T5, a sign of “0” is assigned when the sum S in the X direction is large, and “1” when the sum S in the Y direction is large. "Is assigned (step S2c).

  That is, for each image data in each reference mask area t0, t1,..., T5, when first-order differentiation is performed in each of the X direction and the Y direction, the edge portion of the fingerprint ridge line in each image data is extracted and becomes a numerical value. When the sum S obtained by adding the absolute values in the X direction is large, the fingerprint ridge line flows in the direction of the vertical width b of the rectangular area, and the sum S obtained by adding the absolute values in the Y direction. Is large, it means that the fingerprint ridge line flows in the direction of the lateral width a of the rectangular area. Accordingly, as shown in FIG. 3A, the vertical width b which is the longitudinal direction of the rectangular area. The reference mask areas t0, t2, t3, t4, and t5 in which the fingerprint ridge line flows in the direction of “5” is denoted by “0”, and the reference mask area t1 in which the fingerprint ridge line flows in the width a direction is denoted by “1”. Assigned That.

  Then, in the six reference mask regions t0, t1,..., T5, the reference mask regions t0, t2, to which the code “0” in which the fingerprint ridge line flows in the direction of the vertical width b that is the longitudinal direction of the rectangle is assigned. As shown in FIG. 3B, t3, t4, and t5 are transformed into a horizontal width b [dot] and a vertical width a [dot] (a <b) and converted into vertical and horizontal directions, and are converted into rectangular reference mask regions tn. Rearrangement is made so that as many fingerprint ridges as possible cross (step S2d).

  Then, as shown in FIG. 4A, each of the six reference mask areas t0, t1,..., T5 rearranged on the collation fingerprint image Gs and the registered fingerprint image memory 13b. Each maximum correlation (position) region k0, k1,..., K5 for which the maximum correlation coefficient is calculated on the image Go is searched for each reference mask as shown in FIG. 4B (step S3). , S4).

  Then, the positional relationship XYt between the reference masks (regions) t0, t1,..., T5 arranged in the verification fingerprint image Gs and the maximum correlation (position) regions k0, k1,..., Detected on the registered fingerprint image Go. The positional relationship XYk between k5 is calculated, and the similarity of each positional relationship XYt: XYk is determined (steps S5 and S6).

  Then, an error (positional deviation) between the positional relationship XYt between the reference masks (regions) t0, t1,..., T5 and the positional relationship XYk between the maximum correlation (position) regions k0, k1,. It is determined whether or not the degree of similarity (positional deviation “small” → similarity “high”) is equal to or greater than a preset threshold value (step S7), and the matching fingerprint image Gs is determined to match the registered fingerprint image Go ( Step S8) or a mismatch determination is made (step S9).

  Then, a matching result of the matching fingerprint image Gs with the registered fingerprint image Go is stored in the matching determination result memory 13g. For example, when this fingerprint matching device is used in an entrance / exit management system, Entry / exit is permitted / rejected according to the match / mismatch determination result.

  Therefore, according to the fingerprint collation device of the first embodiment having the above configuration, when collating the collation fingerprint image Gs with the registered fingerprint image Go, the rectangular reference mask regions t0, t1,. , T5 and the relative position distribution between the maximum correlation areas k0, k1,..., K5 on the other fingerprint image where the maximum correlation coefficient is calculated with the image in each area. The identity of each fingerprint image is determined, and when a plurality of rectangular reference mask regions t0, t1,..., T5 are arranged for one fingerprint image (Gs or Go), the rectangular reference masks are arranged. The regions t0, t1,..., T5 are provisionally arranged vertically with respect to a predetermined position serving as the center of gravity of the fingerprint image, and then the edge direction of the fingerprint image in each reference mask region t0, t1,. Analyze the mask area The fingerprint image to be collated is re-arranged in the direction in which the longitudinal direction of the reference mask rectangular area is perpendicular to the direction in which the fingerprint ridge line crosses as many as possible, that is, the direction of the fingerprint ridge line direction. A more characteristic image area on (Gs or Go) can be grasped as a reference mask area t0, t1,..., T5, and by comparing images based on the reference mask areas t0, t1,. High-precision fingerprint verification / authentication processing can be performed.

  In the reference mask arrangement process (step S2) in the fingerprint collation apparatus of the first embodiment, a plurality of rectangular reference mask areas t0, t1,..., Tn arranged on one fingerprint image (Gs or Go) are obtained. After temporarily arranging the center of gravity on the fingerprint image as a reference, the direction of the edge of the fingerprint image for each reference mask area t0, t1,..., Tn is analyzed to selectively select the arrangement direction of the rectangular mask area tn. Although the vertical / horizontal conversion is performed, as described with reference to FIGS. 7 and 8, the fingerprint ridge line direction is constant from among a large number of rectangular regions T00 to Tij arranged on one fingerprint image. A plurality of rectangular reference mask areas t0, t1,..., Tn satisfying the conditions are searched and temporarily arranged, and thereafter, in the same manner as described above, according to the edge direction of the fingerprint image for each reference mask area t0, t1,. The rectangular mask area t Orientation of may be selectively configured to aspect convert.

  FIG. 7 is a flowchart showing the mask arrangement area search process executed in the preceding stage in the reference mask arrangement process (step S2) accompanying the fingerprint collation process by the fingerprint collation apparatus 10.

  FIG. 8 is a diagram showing a search arrangement state of a plurality of rectangular reference mask regions t0, t1,..., Tn on one fingerprint image associated with the mask arrangement area search process in the fingerprint collation apparatus 10.

  That is, as shown in FIG. 8, first, 24 pieces of width a [dot] and length b [dot] (a <b) are compared with the collation fingerprint image data Gs stored in the collation fingerprint image memory 13c. The vertically long rectangular areas are arranged in a matrix of four columns and six columns in the horizontal direction T00 to T35 (step A1), and are arranged to fill the image (step A1), and the individual image data in each rectangular area T00 to T35 are arranged in the X and Y directions. The sum S of absolute values of the primary differential values, which are differences between adjacent pixels in each direction, is calculated (step A2).

  In each of the rectangular regions T00 to T35, a code “0” is assigned when the sum S in the X direction is large, and a code “1” is assigned when the sum S in the Y direction is large (step A3). .

  That is, with respect to the individual image data in each of the rectangular regions T00 to T35, when first-order differentiation is performed in each of the X direction and the Y direction, the edge portion of the fingerprint ridge line in each image data is extracted and becomes a numerical value. When the sum S added in the X direction is large, the fingerprint ridge line flows in the direction of the vertical width b of the rectangular area, and when the sum S obtained by adding the same absolute value in the Y direction is large. The fingerprint ridge line flows in the direction of the lateral width a of the rectangular area. Accordingly, the rectangular area Tij in which the fingerprint ridge line flows in the direction of the longitudinal width b that is the longitudinal direction of the rectangle The code “1” is assigned to the rectangular area Tij in which the fingerprint ridge line flows in the direction of the horizontal width a.

  Then, among the rectangular regions T00 to T35 to which the code “0” or “1” is assigned according to the flow direction of the fingerprint ridge line, six reference mask regions t0, which are two vertical rows × three horizontal rows. With respect to t1,..., t5, the number of areas C (= 0 to 6) to which the code “0” is allocated is counted while being moved one by one in the vertical and horizontal directions as indicated by the arrow Q. The positions of reference mask areas t0, t1,..., T5 where the number of areas C is (C1 <C <C2) are searched (step A4).

  That is, for example, when C1 = 2 and C2 = 5, among the six reference mask regions t0, t1,..., T5, fingerprint ridge lines flow in the vertical direction in the three to four reference mask regions, and the remaining four The three reference mask areas are searched as positions where the fingerprint ridge lines are flowing in the horizontal direction, and thereby the reference where the fingerprint ridge lines are flowing in a good balance between the vertical and horizontal directions in the entire fingerprint image Gs area. The mask areas t0, t1,..., T5 are searched.

  Then, reference mask areas t0, t1,..., T5 are arranged at positions searched from the 24 rectangular areas T00 to T35 in the collated fingerprint image Gs on the condition (C1 <C <C2) (step). A5).

  As a result, six reference mask regions t0, t1,..., T5 can be arranged on the collated fingerprint image Gs at positions where the fingerprint ridge lines flow with good vertical and horizontal balance. .., T5 selective vertical / horizontal conversion processing (step S2) and the maximum correlation areas k0, k1,..., K5 on the registered fingerprint image Go. Search processing (steps S3, S4), similarity calculation processing (steps S5, S6) by comparing the positional relationship between each reference mask region t0, t1,..., T5 and each maximum correlation region k0, k1,. By performing the collation match / mismatch determination process (steps S7 to S9) based on this, it is possible to perform the fingerprint collation / authentication process with higher accuracy.

  The fingerprint collation device described so far has a configuration in which the fingerprint image read by the image reading device 14 is directly input to both the collation fingerprint image Gs and the registered fingerprint image Go, and the fingerprint collation processing in FIG. 5 is performed. However, as will be described with reference to FIG. 9, the fingerprint image reading state is determined by detecting the pressing state of the finger F against the fingerprint reading surface 14 m of the image reading device 14 and improperly read the fingerprint. A configuration may be employed in which images are excluded or the reference mask regions t0 to tn are prohibited from being placed at inappropriate positions.

  FIG. 9 is a diagram showing a configuration of a fingerprint reading surface 14m in the image reading device 14 of the fingerprint collation device.

  That is, the image reading device 14 is configured such that the pressure detection element 14s is provided so as to overlap the back surface of the fingerprint reading surface 14m, and the pressure pressure distribution by the finger F in each part of the read fingerprint image G can be detected.

  Then, when the pressing pressure of the finger F is equal to or less than the threshold value in the area of the read fingerprint image G that is larger than the preset area, it is determined that the finger F is not correctly placed on the fingerprint reading surface 14m. A warning message such as “reading failure” is output, and the user is prompted to reread the fingerprint image G by repositioning the finger F.

  As a result, it is possible to solve the problem of fingerprint registration and fingerprint collation using an inappropriate fingerprint image G read without the finger F being placed correctly.

  Further, based on the pressure pressure distribution by the finger F in each part of the read fingerprint image G, a region where the pressure distortion is large is distinguished from a region where the pressure distortion is not large, and the arrangement of the reference mask regions t0 to tn in the region where the pressure distortion is large is prohibited To do.

  Accordingly, fingerprint collation processing can be performed based on the reference mask regions t0 to tn arranged in a region where the reading distortion of the fingerprint image G is small, and more accurate fingerprint collation / authentication processing can be performed.

  In addition to the fingerprint collation device in the first embodiment, as will be described in the next second embodiment, in the pre-process of the fingerprint collation process, the finger F is not placed on the fingerprint reading surface 14m. The pre-scan image fo (x, y) is acquired, and it is determined whether or not there is noise such as residual fingerprints, scratches, and dust on the fingerprint reading surface 14m, and if there is noise, the read fingerprint image f ( The correct fingerprint image fp (x, y) from which the noise is removed from x, y) may be generated, and then a series of fingerprint collation processing in the first embodiment may be performed.

(Second Embodiment)
FIG. 10 is a flowchart showing an image acquisition process performed as a pre-process of the fingerprint verification process in the fingerprint verification apparatus according to the second embodiment of the present invention.

  First, in the state where it is determined that the finger F is not placed because no pressing pressure is detected by the pressure detection element 14s provided to overlap the fingerprint reading surface 14m of the image reading device 14, the image reading device 14 The pre-scan image fo (x, y) of the fingerprint reading surface 14m on which the internal light source directed to the fingerprint reading surface 14m in 14 is set to an appropriate light amount A and is turned on (step H1) and the finger F is not placed. Imaging is acquired (step H2).

  When it is determined that the finger F is placed on the fingerprint reading surface 14m by detecting the pressing pressure of the finger F by the pressure detecting element 14s (step H3), the image reading device 14 has the inside. The internal light source toward the fingerprint reading surface 14m is set again to an appropriate light amount B and turned on (step H4), and a fingerprint image f (x, y) of the finger F is captured and acquired (step H5).

  Thus, when the pre-scan image fo (x, y) of the fingerprint reading surface 14m by the irradiation of the light amount A is imaged and acquired, and the fingerprint image f (x, y) is acquired and acquired by the irradiation of the light amount B, FIG. 11 to determine whether there is noise such as residual fingerprints, scratches, and dust on the fingerprint reading surface 14m based on the prescan image fo (x, y) on the fingerprint reading surface 14m. The

  FIG. 11 is a flowchart showing an image noise determination process performed in the pre-processing of the fingerprint matching process in the fingerprint matching apparatus according to the second embodiment of the present invention.

  In this image noise determination process, first, a prescan image fo (x, y) of the fingerprint reading surface 14m acquired in the image acquisition process in FIG. 10 is acquired (step I1), and this prescan image fo (x , Y), a plurality of rectangular areas Mk (k = 1, 2,..., N) are arranged so as to cover substantially the entire area (step I2).

  Then, edge extraction by first-order differentiation between adjacent pixels is performed on each image in each rectangular area Mk, and an edge strength P that is the sum of absolute values in the area is calculated (step I3), and this calculation is performed. The number of rectangular areas Mk where the edge strength P is equal to or greater than a preset edge strength threshold Tp is counted (step I4), and it is determined whether or not the counted number is equal to or greater than a certain number threshold Tn (step I4). Step I5).

  That is, when noise such as a residual fingerprint occurs in the pre-scan image fo (x, y) on the fingerprint reading surface 14m, the edge of the image accompanying the noise is extracted and calculated as the edge strength P equal to or higher than the threshold value Tp. When such a rectangular area Mk is equal to or greater than a certain number threshold Tn, it is determined that there is noise such as a residual fingerprint (steps I5 → I6).

  On the other hand, if no noise such as residual fingerprint occurs in the pre-scan image fo (x, y) on the fingerprint reading surface 14m, the edge intensity P equal to or higher than the threshold value Tp is calculated without extracting the edge of the image accompanying the noise. Since the rectangular area Mk in which the edge intensity P is equal to or greater than the threshold value Tp is not counted more than a certain number threshold value Tn, it is determined that there is no noise such as a residual fingerprint (steps I5 → I7).

  When it is determined by the image acquisition process in FIG. 10 and the image noise determination process in FIG. 11 whether noise such as a residual fingerprint exists on the fingerprint reading surface 14m before the fingerprint matching process, the noise in FIG. 12 or FIG. When it is determined that noise exists, the generation and acquisition of a normal fingerprint image fp (x, y) in which noise is removed from the fingerprint image f (x, y) acquired by the image acquisition process is entered. Processing is performed.

  FIG. 12 is a flowchart showing a noise removal process (part 1) performed as a pre-process of the fingerprint collation process in the fingerprint collation apparatus according to the second embodiment of the present invention.

  In this noise removal process (1), first, the prescan image fo (x, y) before placing the finger F on the fingerprint reading surface 14m by the image acquisition process in FIG. 10 and the fingerprint image f (x) where the finger F is placed. , Y) is acquired (step JH).

  Further, it is determined whether or not noise such as residual fingerprint exists on the fingerprint reading surface 14m by the image noise determination processing in FIG. 11 (step JI).

  When it is detected that noise such as a residual fingerprint exists on the fingerprint reading surface 14m (step J1), the pre-scan image fo (x, x, set in step H1 in the image acquisition process is detected. y) A ratio α (= B / A) between the light amount A at the time of acquisition and the light amount B at the time of acquisition of the fingerprint image f (x, y) set in step H4 is obtained (step J2). Is multiplied by the pre-scan image fo (x, y) to obtain a corrected pre-scan image f 1 (x, y) (= fo (x, y) · α that becomes a noise component of the pre-scan image fo (x, y). ) Is generated (step J3).

  Then, the corrected pre-scan image f1 (x, y), which is a noise component of the fingerprint reading surface 14m generated in step J3, is subtracted from the fingerprint image f (x, y) acquired in step JH. A regular fingerprint image fp (x, y) (= f (x, y) -f1 (x, y)) is acquired (step J4).

  On the other hand, when it is not detected in step J1 that noise such as a residual fingerprint is present on the fingerprint reading surface 14m (step J1 (NO)), the fingerprint image fo (x) obtained in step JH is detected. , Y) is obtained as it is as a normal fingerprint image fp (x, y) (= f (x, y)) (step J5).

  According to the noise removal process (part 1) in FIG. 12, the light amount A when the prescan image fo (x, y) is captured on the fingerprint reading surface 14m and the light amount when the fingerprint image f (x, y) is captured. The corrected prescan image f1 (x, y) that becomes a noise component of the prescan image fo (x, y) by multiplying the prescan image fo (x, y) by the ratio α (= B / A) to B ) And subtracting the corrected prescan image f1 (x, y) from the fingerprint image f (x, y), thereby removing a normal fingerprint image fp (x, y) from which a noise image such as a residual fingerprint has been removed. ) Can be obtained.

  FIG. 13 is a flowchart showing a noise removal process (part 2) performed in the pre-process of the fingerprint collation process in the fingerprint collation apparatus according to the second embodiment of the present invention.

  In the noise removal process (2), first, the prescan image fo (x, y) before placing the finger F on the fingerprint reading surface 14m by the image acquisition process in FIG. 10 and the fingerprint image f (x , Y) is acquired (step KH).

  Further, it is determined whether or not noise such as residual fingerprint exists on the fingerprint reading surface 14m by the image noise determination processing in FIG. 11 (step KI).

  When it is detected that noise such as residual fingerprint exists on the fingerprint reading surface 14m (step K1), the pre-scan image fo (x, y) and the fingerprint image in which the noise image exists are detected. Frequency conversion is performed for f (x, y), respectively, and a frequency component go (p, q) of the pre-scan image and a frequency component g (p, q) of the fingerprint image are obtained (step K2).

  Here, a Fourier transform equation for obtaining the frequency component g (p, q) from the fingerprint image f (x, y) is shown in the following equation (1).

  At this time, the signal that appears as the frequency component go (p, q) of the pre-scan image becomes the noise signal component present on the fingerprint reading surface 14m, and appears as the frequency component go (p, q) of the pre-scan image. The noise signal component is removed from the frequency component g (p, q) of the fingerprint image, and the frequency component g (p, q) of the fingerprint image from which the noise signal component is removed is obtained (step K3).

  Then, the frequency component g (p, q) of the fingerprint image from which the noise signal component existing on the fingerprint reading surface 14m is removed is inversely transformed to generate a normal fingerprint image fp (x, y) (step) K4).

  Here, the inverse Fourier transform formula for obtaining the normal fingerprint image fp (x, y) by inverse transform of the frequency component g (x, y) of the fingerprint image is shown in the following formula (2).

  On the other hand, if it is not detected in step K1 that noise such as a residual fingerprint is present on the fingerprint reading surface 14m (step K1 (NO)), the fingerprint image fo (x) obtained in step KH is detected. , Y) is obtained as it is as a regular fingerprint image fp (x, y) (= f (x, y)) (step K5).

  According to the noise removal process (part 2) in FIG. 13, the frequency component go (p, q) of the prescan image is obtained by frequency-converting the prescan image fo (x, y) on the fingerprint reading surface 14m. ) Is removed from the frequency component g (p, q) obtained by frequency-transforming the fingerprint image f (x, y), and the frequency component g () of the fingerprint image from which the noise signal component has been removed. By inversely transforming p, q), a normal fingerprint image fp (x, y) from which noise images such as residual fingerprints are removed can be obtained.

  Therefore, according to the fingerprint collation device of the second embodiment, the prescan image fo (x, y) in a state where the finger F is not placed on the fingerprint reading surface 14m is acquired and remains on the fingerprint reading surface 14m. It is determined whether or not there is noise such as fingerprints, scratches, and dust. If there is noise, a correct fingerprint image fp (x, y) obtained by removing the noise component from the read fingerprint image f (x, y) is obtained. After the generation, by shifting to a series of fingerprint collation processing in the first embodiment, fingerprint collation / authentication processing with higher accuracy and reliability can be performed without being affected by noise such as residual fingerprints. Can do.

  Note that each processing method by the fingerprint collation apparatus described in each of the embodiments, that is, the fingerprint collation process of the first embodiment shown in the flowchart of FIG. 5 and the reference mask arrangement accompanying the fingerprint collation process shown in the flowchart of FIG. Processing, mask arrangement area search processing in the pre-processing of the reference mask arrangement processing shown in the flowchart of FIG. 7, image acquisition processing of the second embodiment in the pre-processing of the fingerprint matching processing shown in the flowchart of FIG. The image noise determination process based on the image acquisition process in the pre-processing of the fingerprint collation process shown in the flowchart of FIG. 12, the image acquisition process and the image noise determination process in the pre-process of the fingerprint collation process shown in the flowcharts of FIGS. Each of the methods such as noise removal processing (part 1) and (part 2) based on this can be executed by a computer. As a program, it can be stored and distributed in an external recording medium 18a such as a memory card (ROM card, RAM card, etc.), magnetic disk (floppy disk, hard disk, etc.), optical disk (CD-ROM, DVD, etc.), semiconductor memory, etc. it can. Various computer terminals having a fingerprint reading function read the program stored in the external recording medium 18a into the storage device 12 by the storage medium reading unit 18, and the operation is controlled by the read program. The fingerprint image reading / collating function described in each of the embodiments can be realized, and the same processing can be executed by the above-described method.

  The program data for realizing each of the above methods can be transmitted on the communication network (Internet) 20 in the form of a program code, and a computer terminal (program server) connected to the communication network (Internet) 20 21), the above-mentioned program data can be fetched from the storage device 22 to realize the above-described fingerprint image reading and collating function.

  Note that the present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the scope of the invention at the stage of implementation. Further, each of the embodiments includes inventions at various stages, and various inventions can be extracted by appropriately combining a plurality of disclosed constituent elements. For example, even if some constituent requirements are deleted from all the constituent requirements shown in each embodiment or some constituent features are combined, the problems described in the column of the problem to be solved by the invention can be solved. When the effects described in the column of the effect of the invention can be obtained, a configuration in which these constituent elements are deleted or combined can be extracted as an invention.

The block diagram which shows the structure of the electronic circuit of the fingerprint collation apparatus which concerns on embodiment of this invention. The figure which shows the data memory ensured in RAM13 of the said fingerprint collation apparatus. The figure which shows the initial arrangement state of the reference | standard mask area | region (template) t0, t1, ..., t5 on the collation fingerprint image Gs accompanying the fingerprint collation process of the said fingerprint collation apparatus, and the arrangement | positioning state after conversion according to fingerprint image ridgeline analysis . The arrangement state of the reference mask areas (templates) t0, t1,..., T5 on the collated fingerprint image Gs accompanying the fingerprint collation processing of the fingerprint collation apparatus and the maximum correlation position (area) k0, k1 on the registered fingerprint image Go. ,..., K5. 7 is a flowchart showing fingerprint collation processing by the fingerprint collation apparatus 10; 5 is a flowchart showing a reference mask arrangement process accompanying a fingerprint collation process by the fingerprint collation apparatus 10; 7 is a flowchart showing a mask arrangement area search process executed in the previous stage in a reference mask arrangement process (step S2) accompanying a fingerprint collation process by the fingerprint collation apparatus 10; The figure which shows the search arrangement | positioning state of several rectangular reference | standard mask area | region t0, t1, ..., tn on one fingerprint image accompanying the mask arrangement | positioning area search process in the said fingerprint collation apparatus 10. FIG. The figure which shows the structure of the fingerprint reading surface 14m in the image reading apparatus 14 of the said fingerprint collation apparatus. The flowchart which shows the image acquisition process implemented by the pre-process of the fingerprint collation process in the fingerprint collation apparatus of 2nd Embodiment of this invention. The flowchart which shows the image noise determination process implemented by the pre-process of the fingerprint collation process in the fingerprint collation apparatus of 2nd Embodiment of this invention. The flowchart which shows the noise removal process (the 1) implemented by the pre-process of the fingerprint collation process in the fingerprint collation apparatus of 2nd Embodiment of this invention. It is a flowchart which shows the noise removal process (the 2) implemented by the pre-process of the fingerprint collation process in the fingerprint collation apparatus of 2nd Embodiment of this invention.

Explanation of symbols

11 ... CPU
12 ... Storage device 13 ... RAM
13a ... Read image (G) memory 13b ... Registered fingerprint image (Go) memory 13c ... Verification fingerprint image (Gs) memory 13d ... Reference mask size / position information (t0, t1, ..., tn) memory 13e ... Maximum correlation position information (K0, k1,..., Kn) Memory 13f ... Collation similarity (R) memory 13g ... Collation determination result memory 14 ... Image reading device 15 ... Display unit 16 ... Input unit 17 ... Bus 18 ... Storage medium reading unit 18a ... External Storage medium 19 ... Transmission control unit 20 ... Communication network (Internet)
21 ... Program server 22 ... Server storage device Go ... Registered fingerprint image data Gs ... Verification fingerprint image data

Claims (6)

A fingerprint collation device that collates fingerprints based on a plurality of rectangular areas arranged in one fingerprint image and a plurality of maximum correlation areas on the other fingerprint image having the maximum correlation with each area image,
A rectangular area arranging means for arranging a plurality of rectangular areas on the one fingerprint image;
A ridge line direction analyzing means for analyzing a ridge line direction of a fingerprint in each of the images in the plurality of rectangular areas arranged by the rectangular area arranging means;
A rectangle that is rearranged so that the longitudinal direction of the rectangular region crosses the ridge line direction of the fingerprint in each rectangular region analyzed by the ridge line direction analyzing unit with respect to the plurality of rectangular regions arranged by the rectangular region arranging unit. a region rearrangement means,
A fingerprint collation device comprising: A rectangular region temporary arrangement means for temporarily arranged by a plurality of rectangular regions in grid form on prior SL one fingerprint image,
The temporary arrangement ridge line direction analyzing means for analyzing the ridge line direction of the fingerprint in each of the plurality of rectangular areas temporarily arranged in a grid pattern by the rectangular area temporary arrangement means,
The rectangular area arranging means is analyzed by the temporary arrangement ridge line direction analyzing means for a predetermined number of rectangular areas among a plurality of rectangular areas temporarily arranged in a grid pattern by the rectangular area temporary arranging means. In each rectangular area, the rectangular area where the ridgeline direction of the fingerprint is the X direction and the rectangular area where the Y direction is the Y direction are searched and arranged almost in half .
The fingerprint collation apparatus according to claim 1, wherein: further,
Pressure distribution detection means for detecting the pressure distribution of the fingerprint pressing pressure corresponding to each part of the one fingerprint image,
The rectangular area arranging unit arranges a plurality of rectangular areas excluding a portion having a large pressure strain based on a fingerprint pressing pressure distribution in each part of one fingerprint image detected by the pressure distribution detecting unit. The fingerprint collation apparatus according to claim 1 or 2 . further,
Prescan image acquisition means for prescanning and acquiring an image of the fingerprint image reading surface;
Noise image presence determination means for determining whether or not a noise image exists in the prescan image of the fingerprint reading surface acquired by the prescan image acquisition means;
A noise image removing means for removing the noise image from the fingerprint image read on the fingerprint reading surface when the noise image presence determining means determines that the noise image is present in the pre-scanned image on the fingerprint reading surface; Prepared,
The fingerprint collation according to any one of claims 1 to 3 , wherein the one fingerprint image and the other fingerprint image are fingerprint images from which noise images have been removed by the noise image removing unit. apparatus. Controls the computer of the fingerprint collation device that collates fingerprints based on a plurality of rectangular areas arranged in one fingerprint image and a plurality of maximum correlation areas on the other fingerprint image having the maximum correlation with each area image A fingerprint collation processing program for
The computer,
A rectangular area arranging means for arranging a plurality of rectangular areas on the one fingerprint image;
A ridge line direction analyzing means for analyzing the ridge line direction of the fingerprint in each of the plurality of rectangular areas arranged by the rectangular area arranging means;
A rectangle that is rearranged so that the longitudinal direction of the rectangular region crosses the ridge line direction of the fingerprint in each rectangular region analyzed by the ridge line direction analyzing unit with respect to the plurality of rectangular regions arranged by the rectangular region arranging unit. region rearrangement means,
A computer-readable fingerprint collation processing program designed to function as a computer. A computer controls a fingerprint collation device that collates fingerprints based on a plurality of rectangular areas arranged in one fingerprint image and a plurality of maximum correlation areas on the other fingerprint image having the maximum correlation with each area image A fingerprint verification method for
A rectangular area arranging step of arranging a plurality of rectangular areas on the one fingerprint image;
A ridge line direction analyzing step for analyzing the ridge line direction of the fingerprint in each of the images in the plurality of rectangular areas arranged by the rectangular area arranging step;
A rectangle that is rearranged so that the longitudinal direction of the rectangular area crosses the ridge line direction of the fingerprint in each rectangular area analyzed by the ridge line direction analyzing step for the plurality of rectangular areas arranged by the rectangular area arranging step. a region rearrangement step,
A fingerprint collation method characterized by comprising:

Images