论文部分内容阅读
在图象边缘提取与优化的基础上建立了一种新的运动图象信源模型，并通过计算机模拟实现了编解码系统。其基本思想是：根据主观视觉特性，以边缘信息为表达运动图象的空间与时间特征的核心。这种模型基编码方法不需要图象的先验知识，模型建立及方案实现较为简单。针对信源模型，引入基于几何约束的图象匹配方法作为运动分析的基础；并建立了广义运动概念、提出并实现了广义运动分析，丰富了帧间预测编码方法。
Based on the image edge extraction and optimization, a new source model of moving image was established, and the codec system was realized by computer simulation. The basic idea is: According to the subjective visual characteristics, using edge information as the core of the spatial and temporal features of moving images. This model base coding method does not require prior knowledge of the image, the model is established and the program is relatively simple. According to the source model, the image matching method based on geometric constraints is introduced as the basis of motion analysis. The concept of generalized motion is established, the generalized motion analysis is proposed and implemented, and the inter prediction coding method is enriched.