一种新的分频图像编码方法

A New Partitionable Frequency Expansion Approach to Image Encoding

为实现图像在空间分割和频率操作的对称展开,树中的所有节点需要频率展开和空间分割.为此,发展了M 可分频率展开方法,并将它与冗余的空间和频率树结合以实现图像的自适应分析与压缩.对比实验表明,图像的对称展 开的编码质量比图像的非对称展开的编码质量好.

In order to implement a symmetric expansion in segmentation and frequency operations for images, all nodes in the trees need both frequency expansion and segmentation. For this reason, M-partitionable frequency expansion is developed,and is combined with the redundant space and frequency tree to realize image adaptive analysis and compression. The comparison experiment shows that the coding quality of a symmetric expansion of images is better than that of an asymmetric expansion of images.