视频监控应用中图像压缩算法的研究

摘要

DCT-based JPEG image compression algorithm is the core technology of video surveillance applications. Picture quality of conventional DCT is guaranteed,but its complexity is greater,which is not conducive to real-time video transmission. Therefore,the main purpose is to reduce the complexity of the algorithm,and to ensure the transmission quality of the image. Based on this feature that image energy is concentrated in the low-frequency portion of the DCT coefficient,propose a partial intercept method which makes the calcula-tion of 8 *8 DCT be significantly reduced,and you can quickly get important information about the image. The results of experiments show that in the case of no significant decrease in PSNR,the algorithm complexity is significantly reduced.%基于DCT的JPEG图像压缩算法是视频监控应用中的核心技术。虽然传统的DCT其画质有保障，但其复杂度较大，不利于视频实时传输。所以，文中主要研究的目的就是降低其算法复杂度，并且保证传输图像的质量。基于图像能量都集中于DCT系数的低频部分这一特性，文中提出局部截取法，可以快速得到图像的重要信息，在没有明显影响图像质量的情况下，使得二维8*8DCT的计算量大幅降低。由实验论证，在PSNR没有明显下降的情况下，算法复杂度明显降低。