Design And Implementation Of NCC Accelerator Based Real-time Target Tracking System

Target tracking is a complex technique based on computer vision, motion control theory and digital image processing, widely used by both military and our daily life. The thesis mainly introduces the (Normalized Cross-correlation Computation)NCC based real time target tracking system. The system is designed by the SoPC method based on32-bit soft core processor Microblaze, accomplish functions including image information acquisition, target recognition&location and motor control. The thesis also discusses the reasonable system architecture and software/hardware partitioning on the system, increases the performance of the target tracking system.Cross-correlation function computation is the heart of the target tracking algorithm, determines the performance of the real time target tracking system. The thesis respectively designs the hardware acceleration circuits of the NCC algorithms in spatial domain and frequency domain, and compares the differences between these two design methods.The system is implemented with FPGA chip. And after accomplishing the algorithm simulation, hardware simulation, software debug and on-board test of the system, the results show that the system meets the actual needs of the real-time target tracking.