Research on Face Recognition System Based on Intelligent Machine Learning Algorithm

This paper introduces a real-time face detection technology based on TMS320C6201. Through the confidential communication between each subsystem, the synchronization of each subsystem is completed, and the real-time face recognition, feature code extraction, and the closest face matching are carried out. Firstly, Grabcut foreground extraction method is used for pre-background segmentation of recognized images to reduce external interference, and then face detection and identification are carried out according to the segmentation effect. A parallel MPI program is developed by transforming the traditional serialization-based face information updating method into a parallel method. This paper applies existing MPI-based methods and existing web-based facial information acquisition methods to improve the efficiency of existing face recognition technologies. It realizes the distributed processing of the update algorithm in the original face recognition system and enhances the ability of the system to process a large amount of data to achieve the purpose of improving the system performance. The experimental results show that the system combining grab cut and Adaboost algorithm can improve the recognition rate and detection rate, and the recognition speed is faster.