Improved single target identification tracking algorithm based on IPSO-BP neural network

• DOI: 10.2478/amns-2024-0336
• Publication Date: 2024-02-27T20:06:53Z
• Authors: Ting Xie
• Keywords: Tracking (education); Position (finance)

Abstract Driven by deep learning techniques in recent years, single target recognition and tracking have developed significantly, but face challenges of real-time accuracy. In this study, an improved IPSO-BP network is formed optimizing three critical aspects the IPSO algorithm: adjusting inertia weight calculation formula, improving factor, creating a new iterative formula for particle updating, which turn combined with BP neural network. After training, paper constructs algorithm higher efficiency. The Algorithm’s performance comprehensively tested through experimental simulation terms real-time, accuracy stability. results show that Algorithm can achieve frame rate (FPS) up to 31 tracking. IOU value as high about 83% some tests. success different scenarios averages approximately 98.50%, position error controlled within 0.7 m, speed 2.75 m/s. This effectively solves problems current technology areas accuracy, showing stability