Depth image intra-frame coding and decoding method and video codec

Abstract

The present invention discloses a depth image intra-frame coding and decoding method and a video codec, wherein the coding method includes: determining an available prediction mode from prediction modes of a depth prediction unit according to preset rules, and the prediction mode is corresponding to the depth prediction unit The prediction mode in the depth model mode 3DMM3 in the three-dimensional high-efficiency video coding 3D-HEVC; For the available prediction mode, calculate the mean square error of the pixel original value and the pixel prediction value of the corresponding said depth prediction unit, obtain the distortion cost; from Among the available prediction modes, the prediction mode with the smallest distortion cost is selected as the optimal prediction mode of the depth prediction unit; the index information of the optimal prediction mode is encoded, and the depth image intra-frame encoding is performed according to the optimal prediction mode. The invention can reduce the complexity of the video decoder and ensure the intra-frame encoding and decoding of the depth image by using the optimal prediction mode.

Description

technical field [0001] The present invention relates to the technical fields of image video encoding and decoding and intra-frame image prediction, in particular to a depth image intra-frame encoding and decoding method and a video encoding and decoding device. Background technique [0002] Depth images generally consist of sharp edges and smooth non-edge objects. However, traditional intra prediction methods based on texture images cannot describe edge information well. Therefore, in the 98th MPEG conference, HHI (Heinrich Hertz Institute, Germany Hertz Institute) proposed an intra prediction method based on a depth model to encode depth images (H.Schwarz, K.Wegner, "Test Model under Consideration for HEVC based 3Dvideo coding, ISO / IEC JTC1 / SC29 / WG11MPEG, Doc.M12350, Nov.2011, Geneva, Switzerland). This method includes a total of 4 intra-frame prediction modes. Among them, DMM3 (Depth Modeling Mode, depth model mode 3) In , each depth (depth) PU (prediction unit) is divid...

AI Technical Summary

Problems solved by technology

[0020] However, the above scheme still needs to be searched in the decoder, and the complexity of the decoder is still high

Moreover, the search for the above scheme in the encoder / decoder is obtained by minimizing the difference between the pixel prediction value and the pixel reconstruction value of the texture image. Therefore, the optimal prediction mode searched may not be the real optimal prediction mode, which may affect the search accuracy of results

Images