Fingerprint recognition algorithm MZFinger5.0

Micro-positive fingerprint recognition algorithm MZFinger5.0 is a fingerprint recognition algorithm with independent intellectual property rights of Guangzhou Weizheng Intelligent Technology Co., Ltd., after years of market testing, the algorithm is safe and reliable, high recognition rate, and good for dry and wet finger recognition.

Micro-positive fingerprint recognition algorithm MZFinger5.0 refers to a series of clear instructions to solve a series of problems, such as fingerprint image preprocessing, data feature extraction, feature matching, fingerprint recognition and so on. This paper gives an overall overview of the micro-positive fingerprint recognition algorithm MZFinger5.0 through three aspects: fingerprint image preprocessing, fingerprint image feature extraction and fingerprint matching.

First, fingerprint image preprocessing: in the process of fingerprint recognition, the newly acquired fingerprint image will be affected by noise, sweat stains, burrs and other factors, so that the image picture is not clear. The purpose of preprocessing is to improve the quality of the input fingerprint image. To improve the accuracy of feature extraction. The position of fingerprint image preprocessing in the whole fingerprint recognition system is like the effect of foundation on the whole house. The quality of the preprocessed image will affect the process of feature extraction and fingerprint matching. This is the first step to deal with in the process of fingerprint recognition. Fingerprint image preprocessing is generally divided into four steps: image segmentation, image filtering, binarization and thinning.

1. Image segmentation. It mainly means that there is a mixture between the original fingerprint image and the background region, which needs to be isolated from the two, which requires preliminary processing of the image according to the gray level, and then normalization and segmentation to eliminate the background region.

two。 Image filtering. This is the core step in the process of fingerprint image preprocessing, mainly through denoising the fingerprint image affected by noise, while repairing and sorting the image to enhance the contrast of the ridge and valley line structure, and further obtain a clearer image.

3. Binarization. After image filtering, the ridge part is enhanced, but the intensity of the ridge is not exactly the same, which is mainly shown in the difference of gray value. The binarization of an image refers to the conversion of a grayscale image (with a grayscale of 255 degrees) into a binary image containing only black and white grayscale, that is, 0 and 1. In this way, the gray value of the ridge tends to be consistent, the image information is compressed, the storage space is saved, and it is beneficial to fingerprint feature extraction and matching.

4. Refine. It refers to the image thinning of the trend, thickness and other features of the fingerprint after binarization, so as to make the fingerprint lines more smooth.

Second, fingerprint image feature extraction: there are many kinds of fingerprint image feature extraction algorithms, including detail feature extraction based on grayscale image, curve-based feature extraction, singularity-based feature extraction, ridge frequency-based feature extraction and so on. Extracting the feature points of the fingerprint image can effectively reduce the pseudo feature points, extract accurate feature points, improve the matching speed and fingerprint recognition performance, and reduce the false recognition rate and rejection rate of the recognition system.

Third, fingerprint matching: fingerprint feature matching is mainly based on the matching of detail eigenvalues. Fingerprint identification is realized by comparing the input fingerprint detail eigenvalues with the stored fingerprint detail eigenvalues, and a critical value needs to be set up when the two are compared. if the matching is greater than this threshold, the fingerprint matches; when the matching is less than the threshold, the fingerprint does not match. Feature matching is the key link of the recognition system, and the quality of the matching algorithm directly affects the performance, speed and efficiency of recognition.

In fingerprint recognition algorithm, fingerprint image preprocessing, feature extraction and fingerprint matching are required from fingerprint input to matching, which is the basic process of fingerprint recognition algorithm, in which there are still a lot of details in each process, which are not explained in detail. This paper only roughly describes the basic steps of micro-positive fingerprint recognition algorithm MZFinger5.0.