Depth video error hiding method based on 3D-HEVC framework

Abstract

The invention relates to a depth video error hiding method based on a 3D-HEVC framework. At present, an effective method for improving the final quality of an image via an error hiding method under the condition that an error occurs in a depth map is lacked. The method comprises the invention comprises the steps of firstly classifying each 64*64 coded unit block in the depth map to divide the blocks into static blocks and moving blocks; then implementing the operation of directly copying a parity block from a reference frame on the static block for hiding the error, and implementing a vector compensation method on the moving block for hiding the error; and at last further optimizing the performance of the error repair block that is poor in performance after error hiding. The vector compensation method specifically comprises the steps of firstly building a movement sharing vector and airspace vector set, then computing an overall parallax vector set and rebuilding a lost block, and at last computing a matching degree of an external boundary. The method provided by the invention is combined with the new standard characteristic and improves the existing error hiding method, and has excellent recovery effect for a network packet loss phenomenon occurring during network transmission of a three-dimensional video.

Description

technical field [0001] The invention belongs to the technical field of video encoding and decoding, and in particular relates to a 3D-HEVC-based error concealment method for depth image transmission errors. Background technique [0002] Error concealment is carried out for video during network transmission. Due to bad transmission factors such as bad transmission environment or interruption of network transmission, packet loss of transmission information may occur during the transmission process. Error concealment is a technology used when the network retransmission mechanism is not implemented in the event of packet loss in transmission information. [0003] In the MVD format video, not only the transmission error of the texture map, but also the transmission error of the depth map must be considered. The depth map plays an important role in the synthesis of virtual viewpoints in the MVD format. Inaccurate depth maps will make the virtual viewpoint Distortions such as blu...

AI Technical Summary

Problems solved by technology

[0003] In the MVD format video, not only the transmission error of the texture map, but also the transmission error of the depth map must be considered. The depth map plays an important role in the synthesis of virtual viewpoints in the MVD format. Inaccurate depth maps will make the virtual viewpoint Distortions such as blurring and ghosting appear in the rendering effect of the image, which in turn affects the experience of the entire 3D video

Images