论文部分内容阅读
在MPEG－2TM5建议的三步自适应量化策略的基础上，提出一种基于人眼视觉特征的MPEG－2视频编码自适应量化算法（PA算法），使MPEG－2的码率分配能基于人眼的特征，即对不同运动速率的物体人眼的细节分解力各不相同。因此可以为不同运动速率的物体分配不同的量化步长，从而提高静止部分图像的分解力，改善了再现图像的主观质量，使之更符合人眼特征。
Based on the proposed three-step adaptive quantization strategy of MPEG-2TM5, an MPEG-2 video coding adaptive quantization algorithm (PA algorithm) based on human visual features is proposed to make MPEG-2 rate assignment based on human The characteristics of the eye, that is, the human eye at different rates of motion of different details of the resolution. Therefore, it is possible to allocate different quantization steps for objects with different motion rates, thereby improving the resolution of the still image and improving the subjective quality of the reproduced image to make it more in line with the human eye characteristics.