CN102819729B - Fingerprint recognizing method - Google Patents

Abstract

The invention discloses a fingerprint recognizing method. The fingerprint recognizing method comprises the following steps: structuring triangles by every characteristic point of a fingerprint to be recognized and a known fingerprint; judging that the triangles are equal to each other; representing every characteristic point of two fingerprints by three ordered arrays formed by the distance from every characteristic point of two fingerprints to three top points of a respective reference triangle; comparing every characteristic point of two fingerprints except for three top points of the reference triangle, and recording the time when three ordered arrays of every characterize point with the same type are equal to each other; at last, according to the equal time of the fingerprint characteristic point, judging the similarity of two fingerprints. The fingerprint recognizing method has the following positive effects that: the recognizing rate of the fingerprint is improved by applying the method based on a triple structure to a fingerprint recognizing system, thereby solving the problems in the existing fingerprint recognizing technology that the situations that the fingerprint image to be matched and the original fingerprint image are displaced, rotated and incomplete are very difficult to correct or exactly position, and the fault-tolerance ability is low.

Description

A fingerprint recognition method

technical field

The invention belongs to the technical field of fingerprint identification, and in particular relates to a fingerprint identification method.

Background technique

The existing fingerprint recognition technology based on feature points first collects the fingerprint image, and then extracts the fingerprint feature points. When identifying, if there are singular points, then use the singular points to align and identify; if there are no singular points, alignment is very difficult, and Larger errors will occur. As a result, the current fingerprint recognition based on feature points is difficult to align and calibrate, resulting in the following shortcomings:

1. When the fingerprint to be compared and the original fingerprint are rotated, it is difficult to calibrate and align the existing identification methods.

2. When the fingerprint to be compared and the original fingerprint are both displaced and rotated, it is more difficult to calibrate and align the existing identification methods.

3. If the fingerprint to be compared is incomplete, the fault tolerance of the existing identification method will be significantly reduced, and this situation is often encountered when the police solve a case, so the existing fingerprint identification method can no longer meet the public security detection. work needs.

Contents of the invention

In order to overcome the above-mentioned shortcomings of the prior art, the present invention provides a fingerprint identification method.

The technical scheme that the present invention adopts to solve its technical problem is: a kind of fingerprint recognition method, comprises the following steps:

Step 1, using an image acquisition instrument to collect all image information of the fingerprint;

Step 2, processing the information of the fingerprint image, and extracting the feature points of the fingerprint;

Step 3, constructing a triangle with each feature point of the fingerprint to be identified and the known fingerprint respectively;

Step 4, judge whether each feature point of the fingerprint to be identified is equal to the triangle constructed in step 3 by each feature point of the known fingerprint:

1) If all are not equal, it means that the two fingerprints are not the same fingerprint, and the comparison ends;

2) If the triangle constructed by a certain characteristic point of the fingerprint to be recognized is equal to the triangle constructed by a certain characteristic point of the known fingerprint, go to step five;

Step 5, taking the triangle determined in step 4 as the reference triangle respectively, and representing each feature point of the two fingerprints with an ordered three-array consisting of the distances from each feature point of the two fingerprints to the three vertices of the respective reference triangle;

Step 6. Compare each feature point in the two fingerprints except for the three vertices of the reference triangle, and record the number of times that the ordered three groups of each feature point of the same type are equal;

Step 7. According to the equal number of fingerprint feature points recorded in step 6, use the existing fingerprint similarity judgment method to judge the similarity of the two fingerprints: if the two fingerprints are the same fingerprint, the judgment ends; if not For the same fingerprint, continue to compare whether the triangle constructed by the remaining feature points of the fingerprint to be identified is equal to the triangle constructed by a certain feature point of the known fingerprint. Triangles are all aligned.

The method for constructing a triangle with feature points described in step 3 is as follows: each fingerprint feature point is used as the criterion, and the two feature points closest to it are found in the fingerprint, and it is judged whether the point and the two closest points form a triangle. If it is a triangle, record the triangle; if not, mark the feature point to fail to construct the triangle, and the feature point will not participate in the comparison, and continue to construct the triangle with other feature points.

Compared with the prior art, the positive effect of the present invention is: by using the method based on the triplet structure in the fingerprint recognition system, the recognition rate of the fingerprint is improved, and it solves the problem that when the matching The fingerprint image and the original fingerprint image are displaced, rotated, and incomplete, which makes it difficult to calibrate or locate accurately, and has low fault tolerance.

Detailed ways

A fingerprint identification method, comprising the steps of:

Step 1, using an image acquisition instrument to collect all image information of the fingerprint;

Step 2, processing the information of the fingerprint image, and extracting the feature points of the fingerprint;

Step 3. Construct a triangle with each feature point of the two fingerprints (the fingerprint to be recognized and the known fingerprint), the specific method is as follows:

Take each fingerprint feature point as the standard, find out the two closest feature points in the fingerprint, judge whether the point and the two closest points form a triangle, if yes, record the triangle; if not, then Mark this feature point to fail to construct a triangle, this feature point does not participate in the comparison, and continue to use other feature points to construct a triangle according to the method described in this step.

Step 4, judge whether each feature point of the fingerprint to be identified is equal to the triangle constructed in step 3 by each feature point of the known fingerprint:

1) If all are not equal, it means that the two fingerprints are not the same fingerprint, and the comparison ends;

2) If the triangle constructed by a certain characteristic point of the fingerprint to be recognized is equal to the triangle constructed by a certain characteristic point of the known fingerprint, go to step five;

Step 5. Take the triangle determined in step 4 as the reference triangle, and use an ordered three-array (a, b, c) composed of the distances from each feature point of the two fingerprints to the three vertices of the respective reference triangle to represent the two Each feature point of a fingerprint:

The principle of this step is as follows:

For a given △ABC on the plane, each point P on the plane determines a unique triad (a, b, c), where a, b, c are the three vertices A from point P to △ABC , the distance of B, C; on the contrary, given any three arrays (x, y, z) may not be able to determine the only point Q on the plane, it takes x, y, z as its vertex A to △ABC, The distance between B and C.

According to this fact, the present invention uses the triangle determined in step 4 as the reference triangle, and the entire fingerprint feature point is not represented by traditional rectangular coordinates or polar coordinates, but is formed by the distance from the feature point to the three vertices of the triangle. Ordinal three arrays (a, b, c) to represent.

This approach was taken because:

1. In the pattern recognition of this construction method, as long as an equal triangle is found in two planes, the vertices of the two triangles can be used as the reference to calculate the distance from the points or feature points on the plane to the plane. The scalar distance of the three vertices of the triangle before matching. In this way, the matching can be very accurate, no calibration is required, and there is no error transmission, and the amount of calculation is small.

2. When matching an incomplete plane or plane feature point with a complete plane or plane feature point, if the traditional matching method is used, this problem is an NP problem, and the amount of calculation is quite large. If this method is used, that is : First find a triangle on the two planes. If the two triangles are equal, you can use the method of 1 to match. This is not an NP problem, which reduces the amount of calculation and is more accurate.

Step 6. Compare each feature point in the two fingerprints except the three vertices of the reference triangle, and record the number of times that the three groups (a, b, c) of each feature point of the same type are equal;

Step 7. According to the equal number of fingerprint feature points recorded in step 6, use the existing fingerprint similarity judgment method to judge the similarity of the two fingerprints: if the two fingerprints are the same fingerprint, the judgment ends; if not For the same fingerprint, continue to compare whether the triangle constructed by the remaining feature points of the fingerprint to be identified is equal to the triangle constructed by a certain feature point of the known fingerprint. Triangles are all aligned.

Claims (1)

1. A fingerprint identification method, characterized in that: comprise the steps: Step 1, using an image acquisition instrument to collect all image information of the fingerprint; Step 2, processing the information of the fingerprint image, and extracting the feature points of the fingerprint; Step 3. Construct a triangle with each feature point of the fingerprint to be recognized and the known fingerprint respectively: The method of constructing a triangle with feature points is as follows: take each fingerprint feature point as the standard, find out the two feature points closest to it in the fingerprint, and judge whether the point and the two closest points form a triangle, if so, then Record the triangle; if not, mark the feature point to fail to construct the triangle, the feature point does not participate in the comparison, and continue to construct the triangle with other feature points; Step 4, judge whether each feature point of the fingerprint to be identified is equal to the triangle constructed in step 3 by each feature point of the known fingerprint: 1) If all are not equal, it means that the two fingerprints are not the same fingerprint, and the comparison ends; 2) If the triangle constructed by a certain characteristic point of the fingerprint to be recognized is equal to the triangle constructed by a certain characteristic point of the known fingerprint, go to step five; Step 5. Take the triangle determined in step 4 as the reference triangle, and form an ordered three-group array composed of the distances from each feature point in the two fingerprints except the three vertices of the reference triangle to the respective three vertices of the reference triangle to represent each feature point of the two fingerprints; Step 6. Compare each feature point in the two fingerprints except for the three vertices of the reference triangle, and record the number of times that the ordered three groups of each feature point of the same type are equal; Step 7. According to the equal number of fingerprint feature points recorded in step 6, use the existing fingerprint similarity judgment method to judge the similarity of the two fingerprints: if the two fingerprints are the same fingerprint, the judgment ends; if not For the same fingerprint, continue to compare whether the triangle constructed by the remaining feature points of the fingerprint to be identified is equal to the triangle constructed by a certain feature point of the known fingerprint. Triangles are all aligned.