Research on real-time infrared image matching acceleration technology based on dual multi-core DSP+FPGA heterogeneous system

This study focuses on enhancing the performance of real-time infrared image matching. Addressing the computational efficiency bottlenecks of traditional methods, a novel acceleration technique based on a heterogeneous system of dual multi-core DSPs and FPGAs is proposed. This scheme leverages the software flexibility of DSPs and the hardware parallelism of FPGAs to construct an efficient real-time image matching platform. The research emphasizes the architectural design of the heterogeneous system, task partitioning strategies, and optimization methods, aiming to accelerate the infrared image matching process, meet real-time requirements, and provide high-performance solutions for related application areas.