Video encoding method, decoding method, device thereof, program thereof, and storage medium containing the program

Abstract

A video encoding method performs video prediction between cameras by using parallax information between a reference camera image which has been already encoded and a camera image to be encoded and corresponding to the reference camera image, so as to create a parallax-compensated image and encode a differential image between the camera image to be encoded and the parallax-compensated image. For each predetermined division unit of the differential image, depending on whether the parallax-compensated image at the same position is present or absent, i.e., the pixel value of the parallax-compensated image is valid or not, a decoded differential image group obtained by decoding a differential image between the already encoded camera image and the parallax-compensated image or a decoded camera image group obtained by decoding the already encoded camera image is set as a reference object.

Description

technical field [0001] The present invention relates to the coding and decoding technology of multi-viewpoint moving images. [0002] This application claims priority based on Japanese Patent Application No. 2006-353628 filed on December 28, 2006, and uses the content thereof here. Background technique [0003] A multi-view video (multi-view video) refers to a plurality of video images in which the same subject and background are photographed by a plurality of cameras. Hereinafter, a video captured by one camera is referred to as a "2-dimensional video", and a group of 2-dimensional video in which the same subject and background are captured is referred to as a multi-view video. [0004] The two-dimensional moving images of the cameras included in the multi-viewpoint moving image are strongly correlated in the time direction. On the other hand, when the cameras are synchronized, the images of the cameras corresponding to the same time are images of the subject and the back...

AI Technical Summary

Problems solved by technology

[0033] However, in this case, since it is necessary to encode information indicating which prediction method is used for each block, even if the prediction residual can be reduced, it is necessary to encode a large amount of additional information, so it cannot Achieving High Coding Efficiency

[0034] In addition, since only one prediction method can be selected for each block, it is not possible to correspond to the case where the presence or absence of occlusion is different for each part of a block.

Images