Face recognition technology for video surveillance integrated with particle swarm optimization algorithm

Abstract

With the rapid development of video surveillance technology, face recognition has become an important security and surveillance tool. To improve the accuracy and applicability of face recognition in video surveillance, this study improved the Inertia Weight (IW) and Learning Factor (LF) based on the Particle Swarm Optimization (PSO) algorithm. Support Vector Machine (SVM) algorithm and Local Binary Mode (LBP) were used to optimize the processing. The results showed that the optimal solution could be obtained after 10 iterations, and the recognition accuracy reached 92.3%. When the number of iterations reached 40, the recognition accuracy inertia weight reached 99.7%. The average operating time of the original PSO algorithm and the optimized PSO algorithm was 26.3 s and 24.7 s, respectively. This shows that the optimization algorithm not only improves the recognition accuracy, but also shortens the operation time, and enhances the convergence performance and robustness to varying degrees. The improved model can improve the recognition rate of video surveillance system, indicating that the optimization algorithm has great application potential in the video surveillance face recognition.