JP2990495B2 - Biometric Recognition Method in Fingerprint Verification - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a biometric recognition method for identifying whether an imprinted fingerprint is a biometric fingerprint or a pseudo-fingerprint in fingerprint collation. Fingerprint verification is to verify the identity by comparing a previously registered fingerprint with a newly imprinted fingerprint. In this case, the registered fingerprint is obtained by binarizing a multi-valued image signal obtained by imaging the imprinted fingerprint, detecting a plurality of feature points, and registering an area of a predetermined size including the feature points, including its coordinate information. It is stored in a file. Then, at the time of fingerprint collation, the multivalued image signal or the binarized image signal obtained by imaging the imprinted fingerprint is compared with the multivalued image signal or the binarized image signal including the feature points read from the registration file, and the predetermined value is compared. If there is a coincidence point in more than a number of regions, it is determined that the collation matches.

[0002]

2. Description of the Related Art FIG. 7 is a flowchart of a conventional fingerprint collation. A fingerprint is imprinted (C1), and a multi-valued image signal obtained by imaging the imprinted fingerprint with a camera is acquired (C2). At this time, the average luminance of the multi-level image signal in a certain area is obtained. If the average luminance exceeds a preset threshold value, it is determined that a fingerprint is imprinted, and the multi-level image signal is binarized (C
3) Perform collation processing with the registered fingerprint (C4), and determine whether the collation is successful (C5).

[0003] A fingerprint has two major feature points, that is, an end point and a branch point of a ridge. In most fingerprint matching devices, a binary image signal of an area of a predetermined size including this feature point portion is provided. Alternatively, the multivalued image signal is stored in the registration file as registration data including the coordinate information of the area. The fingerprint collation process is to binarize a multi-valued image signal of an imprinted fingerprint or as a multi-valued image signal and collate with registered data of a registered fingerprint. If the number of matches is equal to or greater than a predetermined number, the user is confirmed as having passed the verification (C6), and unlocking is performed. If the number is not equal to or more than the predetermined number, it is judged as rejected and the collation processing failure processing is performed (C
7).

FIG. 8 is a flowchart of a conventional fingerprint registration, in which a fingerprint is imprinted (D1) and a multi-valued image is captured (D1).
2), the steps of binarization (D3) are the same as steps (C1) to (C3) at the time of fingerprint collation in FIG. 7, except that the binarized image is thinned and this thinning processing is performed. With respect to the obtained ridge, an end point and a branch point are extracted as feature points, and a region of a predetermined size of a binary image signal or a multi-value image signal is cut out based on the coordinate information of the feature points, and the coordinate information is extracted. At the same time, a registration process for making registration data is performed (D4).

After the registration process, it is determined whether or not the registration has passed (D5). In other words, if a predetermined number of feature points or more cannot be extracted, the number of registered data items is small, so that the data is rejected and registration failure processing (D7) is performed. In that case, for example, a re-registration process is performed. If a predetermined number or more of feature points have been extracted, the dictionary data is registered as a pass (D6).

[0006]

In the general case of registering and collating as a binary image signal in fingerprint registration and fingerprint collation, a finger of a registrant who has performed fingerprint registration is used with silicone rubber or the like. When a pseudo-fingerprint in the form of a fingerprint is imprinted, the average luminance in a certain area of the multi-valued image signal captured by the camera exceeds a predetermined value. Then, a fingerprint collation process is performed. As a result, there is a problem that identity verification is performed by matching. That is, there is a possibility that the fingerprint is collated by imprinting the fingerprint using a pseudo fingerprint instead of the biometric fingerprint, thereby exploiting the fingerprint. An object of the present invention is to prevent abuse caused by pseudo fingerprint imprinting.

[0007]

According to the biometric recognition method for fingerprint collation of the present invention, an image signal obtained by imaging a fingerprint is binarized, and a black pixel represents a valley and a white pixel represents a ridge. Search for black pixels in the fingerprint image of, identify the continuity of the black pixels, determine whether the number of continuous pixels is less than or equal to a first threshold,
It is determined that the sweat gland is present on the ridge when the value is equal to or less than the first threshold, the sweat gland is counted up over one screen, and it is determined whether or not the count value of the sweat gland is equal to or more than the second threshold.
The method includes a step of determining a biometric fingerprint imprint when the value is equal to or larger than the second threshold value, and determines whether the finger is a living finger or a pseudo finger based on sweat glands existing in the ridge.

The method may further include a step of registering the imprinted fingerprint including the coordinates of all sweat glands when it is determined that the fingerprint is a living body fingerprint. That is, fingerprint registration is performed including the coordinates of the sweat glands together with the feature points such as the end points and the branch points, and the fingerprints can be collated at the time of fingerprint collation, thereby preventing misuse of the pseudo finger.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a flowchart of a first embodiment of the present invention, showing a case of fingerprint collation. First, it is determined whether or not a fingerprint is imprinted (A1). As described above, this is determined to be fingerprint imprinting because the average luminance in a certain area of the multi-valued image signal obtained by imaging the imprint fingerprint by the camera exceeds a predetermined value. Next, the multi-level image signal is fetched (A2), binarized (A3), and a black pixel of the binarized binary image signal is searched (A4).

Then, the continuity of the black pixels is checked to determine whether or not the continuity is equal to or smaller than a threshold value (A5). This is because the valley line of the fingerprint becomes a black pixel, the ridge becomes a white pixel, the sweat glands existing on the ridge become black pixels, and the valley line formed by the black pixels has continuity, but the sweat glands are dispersed in a dot-like manner. Therefore, they do not have continuity. Therefore, when the continuity of black pixels is not equal to or less than the first threshold value, it is determined that the pixel is not a sweat gland, and step (A)
Go to 8).

If the continuity of black pixels is equal to or less than the first threshold value, the pixel is regarded as a sweat gland, and is compared with the coordinates of the sweat gland when the coordinates including the coordinates of the sweat gland are registered as registered fingerprints (A6). If the registered sweat gland coordinates and the imprint fingerprint sweat gland coordinates match, the number of sweat glands detected is counted up (A7). In step (A6), if the registered sweat gland coordinates do not match the imprint fingerprint sweat gland coordinates, the process proceeds to step (A8).

In this step (A8), it is determined whether or not all the areas of the binary image have been searched. If the search has not been completed for all the areas, the step (A8) is executed.
When the process proceeds to 4) and the search for all the regions is completed, it is determined whether the number of detected sweat glands is within the determination range, that is, whether the count value of the number of detected sweat glands is equal to or greater than a second threshold value ( A
9). If the number of detected sweat glands is within the determination range, the process proceeds to collation processing (A10), and if not, it is determined that the finger is not a living finger and forcedly terminated (A13).

In the collation process (A10), the collation process is performed on an area including the characteristic point of the fingerprint, as in the conventional example, and it is determined whether or not a predetermined number or more of coincidence areas exist (A1).
1) If there is more than a predetermined number of matching areas, the identity is confirmed (A12), and if there is no more than a predetermined number of matching areas, a matching failure process is performed (A14).

As described above, steps (A4) to (A4)
According to 9), a sweat gland detection process of the imprinted fingerprint is performed, and it is determined that the living finger has been imprinted when the fingerprint image has the sweat gland and the number of the sweat glands is equal to or more than the second threshold value. Is identified and fingerprint matching is performed, so that abuse by a pseudo finger can be prevented.

As shown in FIG. 2, for example, a binary image of a biometric fingerprint includes a valley line having continuity of black pixels, a ridge formed by white pixels, and a plurality of sweat glands formed by black pixels within the ridges. It exists. FIG. 3 shows a binary image obtained by imprinting a pseudo-finger shaped with silicone rubber or the like on this biometric fingerprint. That is, the sweat glands in the pseudo finger are crushed, and therefore, the binary image of the pseudo fingerprint is
As if the noise component had been removed by image processing, no sweat glands were included.

Therefore, as shown in FIG. 4A, a binary image of the valley line 1, the ridge line 2, and the sweat gland 3, which are enlarged and show only a part, is scanned as shown by a dashed line arrow. To mark black pixels. (B) shows a marked black pixel. Then, as shown in (C), the continuity of black pixels is checked using a mask 4 having a predetermined size.

FIG. 4C shows a case of a mask having a size of 3 × 3 pixels, but a mask having a size of about 16 × 16 pixels can be generally used. In this case, the mask is moved and the number of continuous black pixels in the vertical and horizontal directions of the target pixel is checked. For example, in the case of a mask of 16 × 16 pixels,
A group of black pixels in which an arbitrary number of black pixels in the mask are in an isolated state is determined as the sweat gland 3 instead of the valley line 1 and the coordinate value of the sweat gland 3 is held. When a mask having a small number of pixels is used, the number of continuous pixels in the vertical and horizontal directions is checked. If the number of continuous pixels is equal to or smaller than the first threshold, an arbitrary number of black pixels are in an isolated state. Can be determined. Then, in the case of the biometric fingerprint binary image of FIG. 2, for example, a group of black pixels in the mask that is not continuous with the black pixels outside the mask is determined to be a sweat gland 3, and a large number of the sweat glands 3 are detected. However, in the case of the pseudo-fingerprint binary image of FIG. 3, the sweat gland is zero or close to it. Therefore, the detected sweat glands are counted up, and if the count value is equal to or larger than the second threshold value, it is possible to reliably determine that the fingerprint is a biometric fingerprint.

FIG. 5 is a flow chart of the second embodiment of the present invention, showing the case of fingerprint registration. As in the case of the fingerprint collation described above, it is determined whether or not the fingerprint is imprinted (B1), and the imprint is performed. In this case, the multi-valued image is fetched (B2), the multi-valued image is binarized (B3), and the black pixels of the binary image are searched to check the continuity of the black pixels (B4).

Next, it is determined whether or not the continuity of the black pixels is equal to or less than a first threshold (B5). If the continuity of the black pixels is equal to or less than the first threshold, it is determined that the gland is a sweat gland and counted (B6). Then, it is determined whether or not all areas have been searched (B7). When the search is completed, it is determined whether or not the number of sweat glands detected is equal to or greater than a determination threshold (B8). That is, it is determined whether or not the sweat gland count value is equal to or greater than a second threshold value. If the count value is not equal to or greater than the second threshold value, the fingerprint is determined to be a pseudo fingerprint and forced termination is performed (B14). Second
If it is equal to or more than the threshold value, the coordinates of all sweat glands are registered (B9), and the number of detected sweat glands is registered (B10).

Next, similarly to the conventional example, the binary image signal of the area including the feature point is registered as registration data (B1).
1) It is determined whether or not the registration is successful based on whether or not the number of feature points is equal to or more than a predetermined number (B12). If the number is equal to or more than the predetermined number, dictionary data registration is performed (B13). If there is no predetermined number or more, registration failure processing is performed (B15).

The above-mentioned steps (B4) to (B10) show the steps of the sweat gland detection processing.
By determining whether the fingerprint is a biometric fingerprint or a pseudo fingerprint by sweat gland detection, and by registering the coordinates of the detected sweat gland, abuse due to the pseudo fingerprint can be reliably prevented.

FIG. 6 is a functional block diagram of the embodiment of the present invention, in which 11 is a fingerprint imaging unit, 12 is an input operation unit,
Is an image input unit, 14 is a binarization unit, 15 is a black pixel search unit,
Reference numeral 16 denotes a continuity detection unit, 17 denotes a sweat gland determination unit, 18 denotes a collation processing unit, 19 denotes a registration unit, and 20 denotes a collation determination output unit.

The fingerprint image pickup section 11 includes a camera for picking up imprinted fingerprints, and applies a fingerprint picked-up image signal to the image input section 13. The input operation unit 12 includes a numeric keypad and the like, and inputs a fingerprint registration request, a fingerprint collation request, an identification number, and the like. When the average luminance within the predetermined determination exceeds the predetermined value based on the captured image signal from the fingerprint imaging unit 11, the image input unit 13 determines that the fingerprint is imprinted and captures the image. Digitized into 256 gradation display,
It is added to the binarization unit 14.

The image signal binarized by the binarizing unit 14 is applied to a black pixel searching unit 15 for searching and marking black pixels for one screen, and a continuity detecting unit 16 for marking black pixels. In the sweat gland determination unit 17, the continuity of
If the number of continuous black pixels is greater than or equal to the first threshold, it is determined to be a valley line,
If the number of continuous black pixels is equal to or less than the first threshold, it is determined to be a sweat gland,
This sweat gland is counted up, and if the value counted for one screen is equal to or greater than the second threshold, it is determined to be a biometric fingerprint, and if it does not exceed the second threshold, it is determined to be a pseudo fingerprint.

The collation processing unit 18 registers the fingerprint image determined as a biometric fingerprint with the registration unit 19 or compares the fingerprint image registered with the registration unit 19 with the fingerprint image. The process is also forcibly terminated. When the identity is obtained as a result of the collation processing, unlocking processing and the like are performed in the collation determination output unit 20. The functions of the above-described units can be realized using a processor, an image memory, and the like.

[0026]

As described above, according to the present invention, when a fingerprint image contains a large number of sweat glands, it is determined to be a biometric fingerprint, and is distinguished from a pseudo-fingerprint which is shaped using silicone rubber or the like. When it is determined that the fingerprint is a pseudo fingerprint, registration and collation are forcibly terminated, and misuse of the pseudo fingerprint can be prevented, so that there is an advantage that the reliability of the fingerprint collation system can be further improved. Also, if it is determined to be a biometric fingerprint,
By registering the fingerprint including the coordinates of the sweat glands, there is an advantage that the reliability of the identity verification at the time of fingerprint collation can be remarkably improved.

[Brief description of the drawings]

FIG. 1 is a flowchart of a first embodiment of the present invention.

FIG. 2 is an explanatory diagram of a biological fingerprint binary image.

FIG. 3 is an explanatory diagram of a pseudo fingerprint binary image.

FIG. 4 is an explanatory diagram of a sweat gland detection process according to the embodiment of the present invention.

FIG. 5 is a flowchart according to a second embodiment of the present invention.

FIG. 6 is a functional block diagram according to the embodiment of the present invention.

FIG. 7 is a flowchart of a conventional fingerprint collation.

FIG. 8 is a flowchart of a conventional fingerprint registration.

[Explanation of symbols]

 (A1) to (A14) Fingerprint collation step

Claims (2)

(57) [Claims] An image signal obtained by capturing a fingerprint is binarized,
A black pixel in a fingerprint image for one screen showing a valley line by a black pixel and a ridge by a white pixel is searched, the continuity of the black pixel is identified, and it is determined whether or not the number of continuous pixels is equal to or less than a first threshold value. Then, it is determined that the sweat gland is present on the ridge when the value is equal to or less than the first threshold, the sweat gland is counted up over one screen, and it is determined whether the count value of the sweat gland is equal to or more than the second threshold. A biometric recognition method in fingerprint collation, comprising a step of determining a biometric fingerprint imprint when the value is equal to or greater than a threshold value of 2. 2. The biometric recognition method according to claim 1, further comprising the step of registering the imprinted fingerprint including the coordinates of all sweat glands when the biometric fingerprint is imprinted. .