Significance-driven depth image compression method
A technology of depth image and compression method, applied in image coding, image data processing, instruments, etc., can solve the problem of not considering three-dimensional geometric information, and achieve the effect of increasing rendering effect and improving reconstruction accuracy.
Inactive Publication Date: 2015-11-25
BEIHANG UNIV
View PDF2 Cites 3 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
These methods only use the information of the two-dimensional image domain when compressing the depth image, and do not consider the three-dimensional geometric information.
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0015] Below in conjunction with accompanying drawing and example the present invention is described in further detail:
[0016] The implementation process of the invention includes three main steps: grid saliency calculation, sparse representation of depth image and depth reconstruction. figure 1 A schematic diagram of the overall process of the present invention is shown.
[0017] Step 1: Grid saliency calculation:
[0018] For each vertex v of the three-dimensional grid, define ζ(v) as the average curvature of the grid at the vertex; N(v,σ) is the field vertex set whose Euclidean distance of vertex v is σ, and N(v ,σ)={x||x-v||<σ}, x is a grid vertex. Therefore, G(ζ(v),σ) defines the Gaussian-weighted average curvature of vertex v at scale σ, and its calculation formula is as follows:
[0019] G ( ζ ( v ) , σ ) = Σ ...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention relates to a significance-driven depth image compression method comprising the following steps: first, calculating the grid significance value of each vertex of a three-dimensional network model, and storing the calculated significance values to vertex attributes; then, rendering the significance values of the vertexes to texture to generate a two-dimensional grid significance map; next, using the rendered two-dimensional grid significance map to guide parallel Poisson Disk Sampling to generate random pixel sampling points; and finally, fusing a depth-discontinuous edge extracted through a Laplace edge and constructing the sparse representation of a generated depth image. In the reconstruction phase, edge diffusion is carried out through multi-scale double-wave filtering and up-and-down sampling to restore the original image. The invention provides a compression method for depth images acquired by acquiring depth values from Z-Buffer, the sparse representation of a depth image is constructed by fusing random pixel sampling points and a depth-discontinuous edge so as to achieve the purpose of compression, and edge diffusion is carried out through multi-scale bilateral filtering and up-and-down sampling to restore the original image.
Description
technical field [0001] The invention belongs to the field of computer graphic images. Background technique [0002] DepthImageBasedRendering (DIBR) technology has been widely used, including image-based remote rendering systems, image-based interactive 3D roaming, 3D video games, 3DTV and FTV. These applications all use the depth image under the original viewpoint to transform it to an adjacent virtual viewpoint through a three-dimensional image to provide a three-dimensional immersive visual experience. In order to reduce network transmission bandwidth or storage capacity, a large number of depth images and color images need to be compressed. Compression algorithms for color images are quite mature and international standards exist (JPEG-2000 and H.264). However, compression algorithms for depth images are still at a preliminary stage. [0003] The acquisition sources of depth images are divided into three categories: the first category is to obtain high-precision floati...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06T9/00
Inventor 梁晓辉顾敏杰饶木明王晓川
Owner BEIHANG UNIV