An Efficient Fingerprint Extraction and Matching Algorithm for Imaging Device

Aiming at the problem of how to efficiently extract and accurately match the fingerprint information collected by the imaging device, an efficient fingerprint extraction and matching algorithm is proposed. Firstly, the preprocessed fingerprint image is extracted using the trained convolutional neural network (CNN), and the extracted feature vectors are compared using the elastic deformation matching algorithm. The matching degree between the two fingerprint images is evaluated by calculating the similarity and distance between the feature points. The algorithm adjusts the position of the feature points within a certain range to achieve a more accurate matching effect. The simulation results show that under different imaging device types and fingerprint image quality conditions, the average matching accuracy of the proposed algorithm is 94.53%, and the extraction speed is also faster.