摘要
采用DSP和FPGA协同技术设计实现了一个高性能的MPEG 4视频编码器。FPGA模块完成视频采集、YUV分离、数据I/O等功能,而使用DSP专一进行视频压缩编码。针对DSP片内资源特点设计了片内存储器数据分配方案,并根据该方案优化了MPEG 4视频压缩的数据流模式。提出了基于宏块空间复杂度的宏块类型判断算法,有效地降低了视频压缩算法的计算复杂度。测试结果表明,采用MPEG 4视频标准该视频编码器每秒能够压缩39 2帧CIF图像。
Witb the development of video encoding techniques, video compression algorithms become more complicated. A real-time high resolution video encoder cannot be implemented with a single CPU or DSP. A MPEG-4 video encoder is designed and implemented based on coordinated DSP and FPGA techniques. The FP GA module takes the tasks of video acquisition, YUV separation and data I/O functions, while the DSP is dedicated for video compression. The data flow scheme of the MPEG-4 video compression is optimized by milizing the DSP's on-chip memory. A Macro Block (MB) type judging algorithm is proposed based on MB's space complexity. It reduces effectively the computational complexity of the video compression. The experimental re suits indicate thai the MPEG 4 video encoder implementation can encode 39.2 f/s in CIF resolution.
出处
《航空学报》
EI
CAS
CSCD
北大核心
2005年第1期90-93,共4页
Acta Aeronautica et Astronautica Sinica
基金
国家自然科学基金(69974005)资助项目