Implementation Of Video Encoders On DSP And Researches On Video Preprocess

This paper pays attention to the mainstream of international video standards: MPEG-4 Visual and H.264/AVC Recommendation, and makes use of existing encoders on PC platform to realize stable real-time video encoder on Blackfin 533 DSP of Analog Devices Inc.Based on the former achievement of our laboratory, we go ahead with designing more efficient data stream, memory arrangement, DMA transfer configuration, further code optimization, and debugging, not only doubling the encoding speed, which now satisfies the requirements of real-time encoding for 2CIF resolution, but also improving the output image quality of decoder. Right now, a fast and stable MPEG-4 real-time encoder has been brought out. In addition, we try placing some spatial filters at the head of video encoder, which causes the input images becoming a little fuzzier, but achieves certain denoise effect and remarkable bitrate saving, adapted to some low bandwidth applications in special circumstances.To further improve the coding efficiency, this paper continues to take researches on H.264/AVC Recommendation, and implements an H.264 encoder on BF533 DSP based on the existing PC platform encoder x264. With the auto code optimization of DSP developing environment, data Cache, manual code optimization, and modifying and simplifying encoder algorithms, the encoder is able to process the sequence "News" (CIF format) at almost real time speed, as both motion and texture of this sequence are relatively simple. Further optimization is going to be made to increase the speed of the encoder in the future.In addition, the paper takes theory researches on noise reduction problem for surveillance video. According to the characteristics of surveillance video, we propose a novel algorithm based on background modeling, setting up adaptive background models to accomplish foreground segmentation, reducing background noise via model parameters and foreground noise via modified 3D-extention K-nearest neighbor filter. The experiment results show that the algorithm not only increases visual quality and PSNR of the video, but also decreases the output bitrate of video encoder greatly, which also improves the whole performance of the encoder.