A depth map compression method, device, system and storage medium

Abstract

The present application discloses a depth map compression method, device, system and storage medium, which are used to enable the depth map to be directly compressed by a two-dimensional encoding device, and further facilitate network distribution of the compressed depth map. The method includes: obtaining a depth map to be compressed; obtaining all pixels in the depth map; determining a concentrated area of ​​information entropy of all pixels; splitting the depth map according to the concentrated area, wherein all the pixels contained in the concentrated area are The pixels are split into the main image, and all pixels contained outside the concentrated area are split into sub-images; the information corresponding to the main image and the sub-image is input into the two-dimensional encoding device for encoding, so that the output of the two-dimensional encoding device is consistent with the depth The video code stream corresponding to the figure. Using the solution provided in this application, the depth map can be split based on the condition of the concentrated area of ​​information entropy, so that multiple depth maps after splitting can be directly compressed by a two-dimensional encoding device, which is convenient for network distribution.

Description

technical field [0001] A depth map compression method, device, system and storage medium. Background technique [0002] Depth image, also known as distance image, refers to an image with the distance (depth) from the image collector to each point in the scene as the pixel value, which directly reflects the geometry of the visible surface of the scene. In the prior art, the depth map is usually compressed by encoding it into a png format. Although such an encoding method reduces the space size, the png format file is not suitable for network distribution. [0003] With the development of video coding, there are already a large number of two-dimensional coding devices today. These two-dimensional coding devices can provide coding speeds of hundreds of frames per second. The existing transmission protocols provide various excellent load balancing methods for network transmission of video. , if the depth map can be encoded into a video stream transparent to the user through a t...

AI Technical Summary

Problems solved by technology

[0004] However, the current depth map is difficult to be directly compressed by two-dimensional encoding equipment. Taking the depth map provided by Microsoft Camera as an example, the bit depth is 16 bits. The current video protocol has a maximum encoding bit depth of 12 bits, and most Does not provide lossless compression, while depth map encoding generally requires geometric lossless

Images