论文部分内容阅读
本文研究了一种简单有效的DCT自适应图像码方法,它可应用于给定码率的静止图像压缩.本编码方法在进行“无死区”的线性量化时,通过调整图像的每个16*16宏块的量化步距来完成“自适应性”,即:①量化步距随本宏块的活动性而变化.②量化步距同时受码率均衡缓存的反馈控制.在码率为0.4bit/Pel下,对于256*256,8bit/Pel两幅黑白图像的实验结果仍有相当好的图像质量.
This paper studies a simple and effective DCT adaptive image code method, which can be applied to a given bit rate of still image compression. This coding method performs “no-dead zone” linear quantization by adjusting the quantization step size of each 16 × 16 macroblock of the image to complete “adaptive”, that is: ① quantization step with the macroblock activity Sexual change. ② quantization step by bit rate equalization buffer feedback control. At a bit rate of 0.4 bit / Pel, the experimental results for 256 * 256 and 8 bit / Pel black-and-white images still have quite good image quality.