JPS63153687A - Method and apparatus for detecting finger print pattern - Google Patents

Abstract

PURPOSE:To grasp a characteristic of an entire finger print image comparatively simply by applying the line segment string formation to a thinned line image of a picture of a finger print, collecting the distribution of angles of each line segment and comparing the result with contents of a dictionary storing the distribution of angles of line segments for each pattern. CONSTITUTION:A picture signal data for a finger print subjected to line thinning processing by a line thinning circuit 5 is approximated by polygonal lines in a line segment string forming circuit 6 and the result is represented as the synthesis of line segments and stored in a processing result storage register 7. Then an angle from a reference line segment is obtained from the line segment string data by a lie segment angle calculation circuit 8, and the distribution of frequency with respect too angle and the characteristics of the distribution (such as mean value, variance and extreme value) are obtained in an angle distribution register 9 and stored therein. On the other hand, the distribution of frequency and the characteristics of the distribution with respect to the angle data corresponding to the finger print patterns (such as whorl, left loop, right loop and arch) are stored in the angle distribution dictionary 10 and in comparing and collating the contents of the register 9 with the stored data in the dictionary 10 by a comparator 11, the characteristic of the entire finger print image are grasped.

Description

[Detailed Description of the Invention] [Summary] In detecting a fingerprint pattern, the present invention creates a line segment sequence from binary fingerprint shape data, calculates a line segment angle from the line segment sequence, and calculates the line segment angle. This invention relates to a method and apparatus for detecting fingerprint patterns, in which the frequency distribution of the fingerprint pattern is calculated and compared with a dictionary to detect the fingerprint pattern.

[Industrial application field]

The present invention relates to a method and apparatus for detecting a fingerprint pattern from a fingerprint video signal.

[Conventional technology]

Fingerprint recognition is considered an important area of personal identification. As a conventional technique, there is a method of performing verification based on the characteristics of a certain part of a fingerprint image.

However, this method is not necessarily satisfactory because the characteristics of the fingerprint image as a whole are not grasped.

[Problem that the invention seeks to solve]

In view of the above-mentioned problems, an object of the present invention is to convert a thinned image of a fingerprint image into line segments, aggregate the angular distribution of each line segment with respect to a reference line, and store the angular distribution of the line segment sequence in each pattern. The purpose of this invention is to make it possible to grasp the characteristics of the entire fingerprint image using a relatively simple method and device based on the idea of comparing it with a dictionary.

[Means for Solving the Problems] In one embodiment of the present invention, as illustrated in the process shown in FIG. 2, a fingerprint sensor is used to convert a fingerprint into an electrical signal and the signal is binarized. (22) In the fingerprint pattern detection method, the binary data is thinned (23) (see Figure 3 left), a line segment sequence is created from the thinned data by means such as broken line approximation ( 24) Calculating line segment angles from line segment sequence data (see center of Figure 3); Calculating angle distribution from line segment angle data (25)
(See Figure 3, right). Prepare a dictionary that stores angular distribution data for fingerprint patterns in advance, and compare the calculated angular distribution data (for example, average value, variance, extreme value, etc.) with the contents of the dictionary ( 2
6), is characterized by performing the following steps.

In another embodiment of the present invention, as illustrated in FIG. 1, a fingerprint sensor 1 that lifts and scratches a fingerprint image and converts it into an electrical signal, a binarization circuit 4 that binarizes this electrical signal, and a fingerprint pattern detection device equipped with a thinning circuit 5 that thins the binary fingerprint data, a line segment sequence is created from the thinned data (for example, broken line formation).
A line segment side generation circuit 6 to calculate the line segment angle, a line segment angle calculation circuit 8 to calculate the line segment angle from the line segment sequence data (see the center of Figure 3), and a frequency distribution of line segment angles (see the right side of Figure 3) to store data. It is characterized by comprising an angular distribution register 9, a dictionary 10 that stores line segment angular distribution data for fingerprint patterns, and a comparison circuit 11 that compares the contents of the angular distribution register 9 and the contents of the dictionary IO.

[Function] When the method or device of the present invention described above is used, the binarized data obtained from the entire image of a fingerprint is thinned, for example, is approximated by a broken line by creating a line segment sequence, and the data obtained by the broken line approximation is The line segment angle is determined, and the frequency distribution of the line segment angle, such as data representing its characteristics, is also determined. For typical patterns of fingerprints, the frequency distribution of line segment angles and data representing its characteristics (average, variance, etc.) are prepared in advance.
If a dictionary is created by associating data such as extreme values, etc., fingerprint patterns such as a spiral pattern, a left-hand flow pattern, etc. can be detected from this dictionary, the frequency distribution obtained above, and data representing its characteristics.

〔Example〕

A block circuit diagram of an apparatus for performing a fingerprint pattern detection method as an embodiment of the present invention is shown in FIG. 1, and a process flow diagram thereof is shown in FIG.
As shown in the figure.

This device includes a fingerprint sensor 1, a frame memory 2, a noise removal circuit 3, a binarization circuit 4, a line thinning circuit 5, a line segment generation circuit 6, a processing result storage register 7, a line segment angle calculation circuit 8, and an angle distribution circuit. It includes a register 9, an angle distribution dictionary 10, and a comparison circuit 11.

The fingerprint sensor 1 converts the entire image of the fingerprint whose pattern is to be detected into an image signal. The image signal is stored in frame memory 2. The image signal from the frame memory 2 is transferred to the noise removal circuit 3.
It is supplied to the processing result storage register 7 via. The image signal stored in the processing result storage register 7 is sent to the binarization circuit 4
The signal is binarized into a digital signal using a threshold value, and then thinned through the thinning circuit 5 (see the third on the left). For the steps up to this point, conventionally used methods and equipment can be used as they are.

The image signal data of the fingerprint thinned by the thinning circuit 5 is approximated by a broken line in the line segment generation circuit 6 and expressed as a composite of line segments, and the result is stored in the processing result storage register 7. Next, from the line segment sequence data, the line segment angle calculation circuit 8 calculates the angle θ from the reference line as shown in the center of FIG.

From the angle data obtained as described above, the frequency distribution for the angle is obtained in the angle distribution register 9. Also, the characteristics of this distribution (average value, variance, extreme values, etc.) are determined and stored.

On the other hand, as shown in FIG. 4 as the angle distribution dictionary 10,
The frequency distribution (
Figure 4, lower row) and its distribution characteristics (average, variance, extreme values, etc.) are determined and stored, and are stored in the angular distribution register 9.
If the comparison circuit 11 compares and matches the data stored in the dictionary IO with the data stored in the dictionary 10, the characteristics of the entire fingerprint image can be grasped.

Regarding each operation process of the above-mentioned apparatus, the steps after the noise removal process are shown as a flowchart in FIG. A detailed explanation of steps 21 to 27 will be omitted since it overlaps with the previous explanation. Third
The angle versus power distribution on the right side of the figure explains the process of tabulating the angles showing the characteristics of the fingerprint pattern similar to that in FIG.

〔Effect of the invention〕

According to the present invention, the characteristics of the entire fingerprint image can be grasped using a relatively simple method and device, and the pattern of the fingerprint can be appropriately detected.

[Brief explanation of the drawing]

Fig. 1 is a block circuit diagram of an apparatus for performing a fingerprint pattern detection method as an embodiment of the present invention, Fig. 2 is a process flow chart of an embodiment of the present invention, and Fig. 3 explains the process of Fig. 2. figure to do,
FIG. 4 is a diagram explaining the angular distribution of each pattern of a fingerprint. In the figure, ■... Fingerprint sensor, 2... Frame memory, 3... Noise removal circuit, 4... Binarization circuit, 5... Line thinning circuit, 6... Line segment side creation. Circuit, 7... Processing result storage register, 8... Line segment angle calculation circuit, 9... Angle distribution register, 10... Angle distribution dictionary, 11... Comparison circuit, 21... Noise removal , 22... Binarization, 23... Line thinning, 24... Line segment sequence creation, 25... Angle aggregation, 26... Comparison, 27... Dictionary.

Claims (1)

[Claims] 1. Converting a fingerprint into an electrical signal using a fingerprint sensor, binarizing the converted signal (22), thinning the binarized data (23), and detecting the pattern of the fingerprint. In the fingerprint pattern detection method, a line segment sequence is created from the thinned data (24), a line segment angle is calculated from the line segment sequence data, and an angular distribution is calculated from the line segment angle data. (25) Comparing the calculated angular distribution data with the contents of a dictionary storing predetermined angular distribution data for fingerprint patterns (26) A fingerprint pattern characterized by comprising each of the above steps. Detection method. 2. A fingerprint sensor (1) that captures an image of a fingerprint and converts it into an electrical signal, a binarization circuit (4) that binarizes the converted electrical signal, and a thin line that converts the binarized fingerprint data. A fingerprint pattern detection device comprising a thinning circuit (5) that creates a line segment sequence from the thinned data, a line segment sequence creation circuit (6) that creates a line segment sequence from the line segment sequence data; A line segment angle calculation circuit (8) that calculates the line segment angle, an angle distribution register (9) that stores the frequency distribution data of the calculated line segment angle, and a dictionary that stores the line segment angle distribution data for the fingerprint pattern created in advance. (10); and a comparison circuit (11) that compares the contents of the angular distribution register (9) with the contents of the dictionary (10).