基于端叉特征融合的指纹识别算法OACSTPCD

There are various types of biometric recognition,such as face recognition,fingerprint recognition,DNA gene sequencing,etc.In literature,most of the methods used are complex algorithms or deep learning methods,which are insufficient in real-time ef-fectiveness.In this paper,aiming at the characteristics of complex fingerprint patterns,we use image processing technology and the latest OpenCV4 to implement fingerprint recognition algorithms,avoiding complex algorithms and time-consuming computation in deep learning.Firstly,the collected images are preprocessed using image processing related technologies,including clipping,rota-tion,and multiple filtering;Secondly,a feature extraction algorithm for fingerprint cross fusion is constructed to determine the corre-sponding types and angles,and 1 670 key points for fingerprint feature fusion to be identified are plotted;Finally,the fingerprint rec-ognition functionality is tested,and the following conclusions are drawn from the experiments.The average time consumed by the feature extraction of this algorithm is 47.0 ms,and the average matching time is approximately 7.7 ms.Additionally,the accuracy is the highest among various feature extraction algorithms,reaching 93.8%.Therefore,it can be concluded that this algorithm enables fast and accurate identification and matching of fingerprints in the fingerprint database,effectively improving the precision and effi-ciency of fingerprint recognition.