Motion vector estimation method multiview image encoding method multiview image decoding method motion vector estimation device multiview image encoding device multiview image decoding device motion vector estimation program multiview image encoding program and multiview image decoding program

摘要

It is possible to estimate a motion vector with high accuracy even in a situation where a processed image can not be obtained, and to realize efficient multi-view image encoding by simultaneously using inter-camera correlation and temporal correlation in video signal prediction. A point-in-time synthesized image at a time point at which the processed image is photographed is generated from the reference camera image photographed by the camera different from the process camera that captured the processed image included in the multi-view image, in accordance with the setting such as the process camera. A motion vector is estimated by searching for a corresponding area in a reference image photographed by a processing camera using an image signal on a point-in-time synthesized image corresponding to a processing area on the processed image without using the processed image.