JP2000076422A - Finger and palm print stamping device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make distributable a feature point extracting job to finger/palm print stamping devices and to make improvable the job efficiency by extracting the feature points of a ridge pattern of an inputted ridge image and transferring the feature point information to a collating device via a network. SOLUTION: The ridges 201 of finger and palm prints collected from the nationwide police stations (local stations) and the left fingerprint documents 202 collected at the criminal investigation sites undergo the digital processing 203 via a live scanner or a scanner and are transferred to a center via a network 210 as the redge pattern images. At the center, the feature points are extracted 221 and the feature value information are obtained. The collation/ retrieval 222 is carried out to a feature value information data base based on the said feature value information and a collation candidate is selected. A selected candidate image is called out of a redge image data base of finger and palm patterns, etc., and processed 223. The transferred redge image is compared with the called candidate image and the personal identification processing 224 is carried out by confirming visually those compared images. The personal identification result is returned to the transferer side and confirmed 231 there.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a finger / palm imprinting apparatus for digitizing ridges on human skin, such as fingerprints and palm prints, and more particularly to a collation system using feature points of ridges via a network. TECHNICAL FIELD The present invention relates to a high-performance finger / palm-printing device in a chemical technology.

[0002]

2. Description of the Related Art Known examples related to the present invention include the following.

(1) Hideyuki Tamura: Computer Image Processing: Applied Practice Volume 3, Soken Shuppan (1992-10) (2) Shoji Mizuno: Image Signal Combining Apparatus, JP-A-61-1411
No. 89 (3) Shoji Mizuno: Image signal coding and decoding apparatus,
No. 6-14200 (4) Yoshikazu Osuge: Fingerprint image transmission device, JP-A-07-5
Japanese Unexamined Patent Publication No. 7099 Publication (1) introduces a fingerprint image collating device and collating method in a local area network.

[0004] Known example (2) proposes a method of paying attention to whether or not a local portion of an image is a background and efficiently combining encoded images corresponding to respective regions.

[0005] A known example (3) proposes an encoding method of a narrowing process which does not allow image deformation.

In a known example (4), a skeleton pattern of a fingerprint is extracted on the remote side and compressed with 100% resilience so that the image is not degraded even if the center is expanded and the feature is extracted. A compression scheme is proposed.

[0007]

On the local side, ridge images such as finger / palm prints obtained on the local side are merely transferred to the center side.
In the general configuration of a conventional finger / palm-print collation device that centrally manages ridge images and their features at the center, operations such as feature value extraction processing are concentrated on the center side, and the local side that requests the collation processing performs collation. There is a problem that it takes time to obtain a result.

Document 1 does not mention the case where the fingerprint image input unit of the collation device is local and has many fingerprint input units via an open network.

In the known examples (2 to 4), since the fingerprint image is transmitted only by encoding and compressing, the process of detecting feature points (end points and branch points), which is preprocessing for collation, is concentrated at the center. I have.

[0010]

According to the present invention, there is provided a finger / palm-printing apparatus which solves the above-mentioned problems, includes a device for communicating via a network, and a device for electronically imaging ridges on human skin. , A computer that processes the input image, a display device that displays the processing result of the computer, and a pointing device that specifies the display coordinates of the display device, and detects the ridge pattern feature points from the input ridge image. There is provided a means for extracting, and a means for transferring feature point information via a network to a matching device for performing personal matching using feature points.

Further, the apparatus has a device for digitizing a latent finger / palmprint collected from a crime investigation site, and is provided with a means for extracting a ridge pattern-like feature point from a latent fingerprint image. Further, the means for extracting the feature points of the ridge pattern includes means for displaying the ridge pattern and designating the feature points of the ridge with a pointing device.

Further, means for receiving, via a network, a plurality of candidate ridge images obtained by collation and retrieval by the collation device which transferred the ridge image, and displaying the candidate ridge image by comparing the candidate ridge image with the input ridge image. Means for performing

[0013] Further, there is provided a database for registering a latent ridge image and feature point information extracted from the latent ridge image,
Means for searching the database from the feature point information extracted from the imprinted ridge image, means for comparing and displaying the indented ridge image obtained as a result of the database search and the imprinted ridge image, Means for deleting the ridge image and the feature point information obtained as a result of the above from the database.

[0014]

Embodiments of the present invention will be described below.

First, the configuration of the network type finger / palm-printing apparatus of the present invention will be described in section (1), the outline of the collation processing of the present invention will be described in section (2), and the outline of the registration processing will be described in section (3).
This will be described in detail in (4) to (8).

(1) Apparatus Configuration First, the apparatus configuration of the finger / palm-printing apparatus of the present invention is shown in FIG.

A finger / palm-print image input device (live scanner) 101 for directly digitizing ridge patterns on the human body such as fingerprints and palm prints is connected to a computer 102. Further, a device (scanner) 103 for digitizing ridge patterns such as latent fingerprints and palm prints collected from crime scenes is also connected to the computer 102.

The computer 102 includes a live scanner or a device (display) 104 for displaying ridge images and processing results input from the scanner, a pointing device 105 such as a mouse for designating coordinates on the display, and a network 110. A communication port 106 for transmitting and receiving data is connected.

In the network 110, a server computer 122 having a database 123 of ridge pattern feature amount information such as fingerprints and palm prints and a database 124 of ridge pattern images such as fingerprints and palm prints via a communication port 121. It is connected.

(2) Verification Processing In a general personal identification system used for conventional facility management and the like, personal identification information is centrally managed. At the access point, only personal identification information is input, and authentication is performed on the center side. I am concentrated. Similarly, FIG. 2 shows an outline of the collation processing of a configuration in which a live scanner or a scanner is added to a police station in each prefecture and connected to a network in a collation device having a finger / palm-print database centrally integrated.

Ridges 201 such as finger prints collected at police stations (local) nationwide, and latent fingerprint documents 202 collected at the investigation site are ridge pattern images digitized 203 by a live scanner or scanner. Is transferred to the center side via the network 210.

On the center side, feature points (end points of ridges, branch points, etc.) are extracted from the transferred ridge image.
1 and obtain feature amount information.

Based on the feature amount information, the feature amount information database is subjected to collation search 223 to narrow down collation candidates. The narrowed-down candidate images are called 223 from a ridge image database such as a finger / palm-print. The transferred ridge image and the called candidate image are compared and displayed, and individual identification processing 224 is performed by visual confirmation. This personal identification result is returned to the transfer source, and the result is confirmed (231).

In the configuration shown in FIG. 2, for a finger / palm-print image collected at a police station nationwide, the feature extraction work on the center side and the visual check work are concentrated at the center, and it takes time to obtain a collation result. .

Next, FIG. 3 shows an example of a processing outline when the finger / palm-printing and imprinting apparatus of the present invention is introduced into police stations nationwide.

A ridge 301 such as a finger print printed at a police station (local side) nationwide, and a latent fingerprint document 302 collected at an investigation site are digitized 303 by a live scanner or a scanner.

Extraction of characteristic points (edges of ridges, branch points, etc.) from digitally processed ridge images 304
Then, feature amount information is obtained.

The extracted feature amount information and the ridge image are transferred to the center via the network 310.

The center performs collation search 322 in the feature amount database based on the transferred feature amount information to narrow down collation candidates. The narrowed-down candidate image is subjected to a calling process 323 from the ridge image database.

The transferred ridge image and the called candidate image are compared and displayed, and individual identification processing 324 is performed by visual confirmation.

This personal identification result is returned to the verification request source,
The result is confirmed by the finger / palm-stamping device of each prefectural police station that received the result (331).

FIG. 4 shows an example in which individual identification processing by visual confirmation is distributed to police stations in each prefecture.

As in FIG. 3, ridges 401 such as finger prints collected at police stations nationwide and latent fingerprint documents 402 collected at the investigation site are digitized by a live scanner or scanner.
03.

Extraction processing of feature points (end points of ridges, branch points, etc.) from digitally processed ridge images 404
Then, feature amount information is obtained.

The extracted feature amount information and the ridge image are transferred to the center via the network 410.

The center performs collation search 422 in the feature amount database based on the transferred feature amount information, and narrows down collation candidates.

The narrowed candidate image is called 423 from the ridge image database.

The narrowed candidate images are returned to the collation request source via the network 410, the received ridge image and the candidate image are compared and displayed, and individual identification processing 331 is performed by visual confirmation at each prefectural police station.

The personal identification result is confirmed (332).

As described above, in the collation processing of ridge images such as finger / palm-prints according to the present invention, the characteristic amount extraction work is not concentrated in the center, but is distributed to local police stations, so that the personal collation time can be reduced. .

Further, personal identification work by visual confirmation is also required.
Since it is not concentrated in the center but distributed to local police stations, personal verification time can be reduced.

(3) Registration Processing Registration of ridge images such as finger / palm prints when a live scanner or a scanner is added to a police station in each prefecture and connected to a network in the finger / palm print database integrated in the center of FIG. An outline of the process will be described.

The ridges 501 such as finger prints collected at police stations (local) nationwide and the latent fingerprint documents 502 collected at the investigation site are ridge pattern images digitized 503 by a live scanner or scanner. Is transferred to the center via the network 510.

On the center side, feature points (edge points of ridges, branch points, etc.) are extracted 52 from the transferred ridge image.
1 and obtain feature amount information.

The feature point information is registered 523 in the feature amount information database.

The transferred ridge image is registered 524 in the ridge image database.

In the configuration shown in FIG. 5, for finger / palm-print images collected at police stations nationwide, the feature extraction work is concentrated on the center side, and it takes time to complete registration.

FIG. 6 shows an outline of a registration process of a ridge image such as a finger / palm-print when the finger / palm-printing apparatus of the present invention is introduced.

The ridges 501 such as finger prints collected at police stations nationwide and the latent fingerprint documents 502 collected at the investigation site are digitized 503 by a live scanner or scanner.

Extraction of feature points (edges of ridges, branch points, etc.) from digitally processed ridge images 304
Then, feature amount information is obtained.

The extracted feature amount information and the ridge image are transferred to the center via the network 310.

On the center side, the transferred ridge image is registered in the ridge image database 624, and the transferred feature information is registered 623 in the feature database.

As described above, in the ridge image registration processing of the present invention, since the feature quantity extraction processing is not concentrated on the center side and distributed to police stations nationwide, the registration processing time can be shortened. Can be.

(5) Ridge Pattern Features Ridge patterns, such as fingerprints and palm prints, on the skin of a human body are unique to each person and have a feature that remains unchanged throughout their lives.

The ridge pattern on the skin of the human body, such as a fingerprint or a palm print, has a feature point at an end point of a ridge called a minutia and a branch point, and identification of a person is made using these features. Is being done. FIG. 7 shows examples of these feature points.

The branch point of the ridge pattern is indicated by a diamond 701, and the end point is indicated by a circle 711.

FIG. 8 shows an example of feature information of a ridge pattern feature point.

The ID number 801 of the feature point of the ridge image, the type 802 indicating the end point or the branch point of the feature point, the x coordinate 803 indicating the position information of the feature point, the y coordinate 804, the direction 805 of the feature point,
And other information 806 obtained from an image and an interrelation between feature points using the above information.

FIG. 9 shows feature point orientation information.

The direction 902 opposite to the ridge direction following the end point 901 of the feature point and the direction 912 opposite to the direction of the inner space formed by branching the branch point 911 of the feature point are defined as the direction information of the feature point.

(6) Means for extracting feature points FIG. 1 shows an example of an algorithm for extracting feature points in a ridge image.
0 is shown.

Step 1010: A ridge image is read from a live scanner or a scanner or a storage medium.

Step 1020: Binarize by threshold processing to separate the ridge region from the other region.

Step 1030: A thinning process is performed.

Step 1040: A binary mask of a feature point is prepared, and the feature point is extracted by mask processing.

Step 1050: The false feature point is removed based on the appearance knowledge information of the false feature point.

Step 1060: The points extracted in the above processing are set as feature points.

In the above processing, as the appearance knowledge of the false feature point, the feature point generated by a ridge cut by a wrinkle or the like forms two pairs, and the false feature point is deleted.

FIG. 11 shows an example of the feature point designating means.

A pointing device 1102 for operating an arrow-shaped icon displayed on the display 1101 is provided.

A feature point designation window 1101 on the display includes a region 1111 for displaying a ridge image, a button 1112 for selecting an enlarged display of the displayed image, a button 1113 for selecting a reduced image, and changing a point designation mode on the screen. Button 1114,
It comprises a button 1115 for calling the information screen of the ridge image.

The point specification mode on the screen includes three modes: end point specification, branch point specification, and feature point deletion.

The branch point 1121 specified in the display area is marked by a diamond, and the end point 1122 is marked by a circle.

FIG. 12 shows a ridge image information display screen.

A ridge image information display window 1201 is a selection list for selecting a sampling method from an electronic method, an ink method, and a resident.
1202, right, left, part? (Unknown), and select finger, palm, sole,? (Unknown) and if it is a finger, thumb, forefinger, middle finger, ring finger, little finger,?
(Unknown), and if it is a palm, palm print, palm, thumb, four finger plane,? Selection list 1203 to select any of (unknown), and selection list to select fingerprint classification to any of spiral print, right footprint, left footprint, bow print, and others
1204, button 12 to call up the screen to reference and display the ridge image
11, a button 1212 for calling a screen for inputting / referencing and displaying ID information, a button 1213 for calling a screen for displaying feature information as shown in FIG. 8, and a ridge image and feature information via a network to the matching device. A button 1221 for transferring and requesting a collation search, a button for transmitting a ridge image and feature amount information to a collation device via a network and requesting a registration process
1222, button to close ridge image information display window 1201
It consists of 1215.

For example, when entering ridge information that has been left behind,
In the site selection, an unknown is selected as appropriate, and a collation search is requested.

By providing the above-described feature point extracting means in the finger / palm-stamping devices of police stations nationwide, it is possible to distribute the feature point extraction work concentrated on the center side.

(7) Image Comparison Means FIG. 13 shows an example of a ridge image comparison means of a candidate image collated and searched for by the collation device on the center side.

A ridge image comparison window 1301 includes a sub window 1302 for displaying an image obtained by a local live scanner or a scanner, and a sub window 1303 for displaying a plurality of ridge images obtained by collation search at the center. A button 1307 for instructing enlarged display of the ridge image, a button 1308 for instructing reduced display, and a button 1305 for closing the ridge image comparison window are provided.

A sub window 1302 for displaying an image selected by the local live scanner or the scanner includes a button 1321 for instructing reading of an image from the live scanner or the scanner or a storage medium, an enlarged display area of a ridge image.
1322, the entire display area 1323 of the ridge image.

The position of the entire enlarged display area is indicated by a rectangular area 1324. In the enlarged display area, a scroll bar 1325 is used to move the display area in parallel.

A sub-window 1303 for displaying a plurality of ridge images obtained by the collation search on the center side includes a button 1331 for designating a screen call for displaying various information (ID information, feature information, etc.) of the search image, A button 1337 for designating switching to the candidate image, an enlarged display area 1332 of the ridge image, and an entire display area 1333 of the ridge image are provided.

The position of the entire enlarged display area is indicated by a rectangular area 1334.

In the enlarged display area, a scroll bar 1335 is used to move the display area in parallel.

By having the image comparing means described above,
The task of confirming personal identification by visually checking the collation search results concentrated in the center can be distributed to police stations nationwide, and as a result, the collation time can be reduced.

(8) Shaking-off of Fingerprints of Related Persons A ridge image of a surviving ridge obtained from a criminal investigation site contains a large number of ridge images of a person who lives at a site unrelated to a crime. A means for shaking off the concerned ridge image and extracting only the fingerprint of the suspect will be described.

A ridge feature amount database is temporarily created, and the ridge image of the surviving ridge is registered in this database. Then, the ridge image is searched for by the feature amount of the ridge image of the person concerned, and the searched image is deleted from the database. Applying to the ridge images of all the concerned persons, the remaining ridges in the database are used as the ridge images of the suspect.

FIG. 15 is a flowchart showing registration of a ridge image.

Step 1410: A ridge image is input.

Step 1420: The feature points are extracted from the ridge image of the latent ridge inputted in step 1410 to obtain feature amount information.

Step 1430: The feature amount information obtained in step 1420 is registered in the temporary feature amount database 1440.

FIG. 16 is a flow chart of search / deletion using an embossed ridge image.

Step 1510: The ridge image of the person concerned is pressed and input.

Step 1520: Feature points are extracted from the embossed ridge image input in step 1510, and feature amount information is obtained.

Step 1530: The temporary feature database 1540 in which the ridge image is registered is searched using the feature information obtained in step 1520.

Step 1550: As a result of the search in step 1530, if there is no candidate image, the process returns to step 1510;
Continue to 1560.

Step 1560: The ridge image input in step 1510 is compared with the candidate image obtained as a result of the search in step 1530 as shown in FIG.

Step 1570: As a result of step 1560, if no identification is made, the flow returns to step 1510. If identified, go to step 1580.

Step 1580: The data of the candidate image identified in step 1570 is stored in the temporary feature amount database 15
Remove from 40.

As described above, in the processing flow, the ridge image of the person concerned at the site can be shaken off from the ridge image. In addition, since the ridge image impressed by the site personnel is not left in the non-volatile storage medium, it is possible to reduce the resistance of the site personnel.

[0101]

As described in detail above, according to the present invention, the feature point extraction work concentrated on the collating apparatus can be distributed to the finger / palm-printing apparatus, and the work efficiency can be improved. Is played.

[Brief description of the drawings]

FIG. 1 is a diagram showing an example of the configuration of a finger palm printing device according to the present invention.

FIG. 2 is an example of an outline of a collation process when a finger / palm-printing device is introduced into a current collation system;

FIG. 3 is a diagram illustrating an example of an outline of a collation process when a finger / palm-printing device having a ridge feature point extracting function according to the present invention is introduced.

FIG. 4 is a diagram showing an example of an outline of a collation process when a finger / palm-printing device having an image comparison function of collation search results according to the present invention is introduced.

FIG. 5 is a diagram showing an example of an outline of a finger / palm-print registration process when a finger / palm-print imprinting device is introduced into a current collation system;

FIG. 6 is a diagram showing an example of an outline of a registration process when a finger / palm-printing device having a ridge feature point extracting function of the present invention is introduced.

FIG. 7 is an explanatory diagram of feature points of a ridge pattern such as a fingerprint and a palm pattern.

FIG. 8 is a diagram illustrating an example of feature point information of a ridge image such as a fingerprint or a palm print.

FIG. 9 is an explanatory diagram of direction information of feature points (end points, branch points) of a ridge.

FIG. 10 is a flowchart of an example of an algorithm for extracting feature points (end points, branch points) of a ridge according to the present invention.

FIG. 11 is a diagram showing an example of a ridge feature point extracting unit of the present invention.

FIG. 12 is a diagram showing an example of a ridge line information display screen of the present invention.

FIG. 13 is a diagram illustrating an example of a comparison unit with a comparison candidate image according to the present invention.

FIG. 14 is a flowchart of a process of registering a latent ridge image in a temporary feature amount database according to the present invention.

FIG. 15 is a flowchart of a search / deletion process of a temporary feature amount database based on an embossed ridge image according to the present invention.

[Explanation of symbols]

101 ... finger / palm-print image input device (live scanner), 102
… Calculator, 103… Legacy fingerprint reader (scanner),
104 display, 105 pointing device, 106 communication port, 110 network
121: communication port, 122: server computer, 123: ridge feature amount database, 124: image database.

 ──────────────────────────────────────────────────続 き Continuing from the front page (72) Inventor Yoichi Seto 1099 Ozenji Temple, Aso-ku, Kawasaki City, Kanagawa Prefecture Inside Hitachi, Ltd. Systems Development Laboratory Co., Ltd. (72) Inventor Morikami Kuroda 3-chome, Sachicho, Hitachi City, Ibaraki Prefecture No. 1 Within Hitachi Engineering Co., Ltd. (72) Inventor Masaji Aoki 890 Kashimada, Saiwai-ku, Kawasaki-shi, Kanagawa Prefecture F-term in the Information Systems Division, Hitachi, Ltd. 4C038 FF01 FF03 FF05 FG00 5B047 AA25

Claims (5)

[Claims] 1. An apparatus for communicating via a network, an apparatus for electronically imaging a ridge on human skin, a computer for processing an input ridge image, and displaying a processing result of the computer. In a finger / palm-printing device having a pointing device for designating display coordinates of the display device, means for extracting a ridge pattern-like feature point from the input ridge image, and using the feature point A finger / palm imprinting device having means for transferring the feature point information to a matching device for personal matching via a network. 2. The finger / palm-stamping and stamping apparatus according to claim 1, further comprising a device for digitizing a ridge image of a ling taken from a criminal investigation site, and extracting a ridge pattern-like feature point from the ridge image of the ling. And a finger / palm-printing device. 3. A finger / palm print according to claim 1, further comprising means for displaying the ridge pattern and designating the ridge characteristic point by a pointing device. Imprinting equipment. 4. The finger / palm-stamping and stamping apparatus according to claim 1, wherein a plurality of candidate ridge images obtained by collation and retrieval by the collation apparatus are received via a network; A finger / palm-printing device comprising a means for comparing and displaying ridge images. 5. The finger / palm-printing and stamping apparatus according to claim 2, further comprising a database for registering a latent ridge image and feature point information extracted from the latent ridge image, wherein the feature points extracted from the stamped ridge image are registered. Means for searching the database from the information, means for comparing and displaying the ridge image of the latent ridge obtained from the database search, and the ridge image of the latent ridge obtained as a result of the database search and its characteristics Means for deleting point information from the database.