基于HXDSP1042的JPEG压缩算法实现及优化

Implementation and optimization of JPEG compression algorithm based on HXDSP1042

为满足合成孔径雷达(Synthetic Aperture Radar,SAR)系统对SAR原始灰度图处理的实时性要求,在分析了JPEG编码原理和双核HXDSP1042硬件架构的基础上,提出了一种基于HXDSP1042的JPEG压缩算法实现及优化方法。该方法运用HXDPS1042的双核资源及并行汇编指令来实现和优化JPEG压缩算法。实验结果表明,该方法在SAR灰度图大小为512×512及1024×1024时,双核HXDSP1042优化实现的JPEG压缩算法运算时间分别仅为17.36 ms和52.76 ms,有效地缩减了JPEG压缩算法的运行时间,满足合成孔径雷达系统的实时性要求。

In order to meet the real-time requirements of synthetic aperture radar(Synthetic Aperture Radar,SAR)system for SAR raw grayscale image processing,based on the analysis of the JPEG encoding principle and the dual-core HXDSP1042 hardware architecture,a JPEG compression algorithm implementation and optimization method based on HXDSP1042 is proposed.This method uses HXDPS1042 dual-core resources and parallel assembly instructions to implement and optimize the JPEG compression algorithm. The experimental results show that when the SAR gray scale image size is 512×512 and 1024×1024, the JPEG compression algorithm operation time the dual -core HXDSP1042 optimized implementation is only 17.36 ms and 52.76 ms, which effectively reduces the JPEG compression algorithm. The running time meets the real-time requirements of the synthetic aperture radar system.