Average period of images in fingerprint recognition for capacitive fingerprint scanners

The fingerprint recognition technology of Capacitive Fingerprint Scanner, the method of calculating the average texture period of a certain image block is described in some previous fingerprint identification Access Control, the average period of the fingerprint image can be calculated as the average value of the texture period of the foreground image block, but When calculating the average period of a specific image block, the size of the projection window is much larger than that of the image block, and the grain direction may change greatly in the projection window, thus affecting the correct calculation of the average period. On the other hand, the size of the projection window is much larger Due to the size of the image block, the image block will be repeatedly projected, which will also affect the correct calculation of the averaging period.

In view of this, this section first divides the fingerprint image JA6 into disjoint rectangular areas with large size, and calculates the local period information of each area P, then calculates the average period according to the local period information, and calculates the basic method of local period information. Yes, first calculate a curve passing through the center of the area, so that the tangent direction of each point (, y) on the curve is perpendicular to o (, 7), and only consider the curve segment on the curve that is located in the area W, in the previous In some experiments, the calculation method of the average period of fingerprints and the classification method based on the average period are given. The experimental results show that the system penetration coefficient based on the average period is smaller than the ideal system penetration coefficient of the H-call mode. The advantage is that the average period can be calculated for any fingerprint and will not be rejected for classification, while in Henv mode, images without core points will be rejected for classification. Another advantage is that the average period is used for classification between classes. The distance can be measured, and the difference between the average periods can be used as the distance, so that when searching for a fingerprint matching the input fingerprint in the database, the search can be started from the class with the smallest distance from the input fingerprint, and the class and the class in the H, He mode The distance between them is not measurable, and the third benefit is that the average period can be easily combined with other features for fingerprint classification.