Embedded zerotree coding method on basis of inverse Loop subdivision
A technology of embedded zero tree and coding method, applied in image coding, image data processing, instruments, etc., can solve problems such as low compression efficiency
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Publication Date
- 2013-05-01
- Estimated Expiration
- Not applicable · inactive patent
Smart Images
Figure 1 Figure 2 Figure 3
Abstract
Description
technical field
[0001] The invention relates to the technical field of digital media, in particular to a coding method for computer three-dimensional graphics. Background technique
[0002] Two-dimensional image compression has been widely used in applications, among which embedded zerotree coding is one of the effective methods. Image embedded zero-tree coding includes three processes:
[0003] (1) Zero tree prediction, encoding important images with zero tree structure, successive approximation quantization. The wavelet-transformed image forms a tree structure according to its frequency band from low to high. The root of the tree is the node of the lowest frequency sub-band. It has three children located in the corresponding positions of the three sub-low frequency sub-bands. See figure 1 upper left corner. The nodes of the remaining subbands (except the highest frequency subband) have four children located in the corresponding positions of the higher-level subband (due...
Examples
Embodiment Construction
[0037] The present invention will be further described below in conjunction with the accompanying drawings.
[0038] refer to Figure 1 ~ Figure 4, an embedded zero-tree encoding method based on inverse Loop subdivision, the Loop subdivision surface may simplify and decompose the triangular mesh through inverse transformation to generate a progressive mesh (see patent for details: A progressive network based on inverse Loop subdivision Grid generation method, patent number: ZL2006101241528), the vertices and triangles contained in the base grid are greatly reduced, but every time an edge is deleted, an offset information is generated, so the offset information occupies a large storage space. Since the simplified model does not have a large mutation between adjacent points in the triangular mesh, and there is a good correlation between adjacent points, the offsets generated by the simplified prediction process are often small in value. Most of them have the characteristics of ...