一种基于FPGA的互相关函数加速器设计

An FPGA Based Cross-Correlation Accelerator Design

为了实现实时高性能目标识别,设计了一种基于FPGA的互相关函数加速器。通过合理的硬件设计与逻辑复用,使互相关函数计算速度得到提升。实验结果表明,加速器可以高速完成互相关函数计算,满足实时高分辨率图像目标识别系统的需求。

The computation of cross-correlation functions is the heart of the target recognition algorithm,due to the performance of the embedded system,the system cannot achieve the high performance cross-correlation function computation.To realize the real time high performance target recognition,we design an FPGA based cross-correlation function accelerator,with the suitable hardware design and the logic reuse,improve the speed of the cross-correlation function computation.The experiment results show that the accelerator can complete the computation of the cross-correlation function in high speed,and can meet the demands of the real time high definition target recognition system.