MULTI-VIEW MATCHING METHOD AND DEVICE USING FOREGROUND/BACKGROUND SEPARATION

摘要

A multi-view matching method and apparatus for using object or background separation are provided to improve depth discontinuous boundary detection performance by using depth information with color information in object or background separation. The multi-view image received from a multi-view camera and the depth information of a reference viewpoint image received from a depth camera are stored(S101). The object or background area separation of the reference viewpoint image is performed by using the reference viewpoint image and the depth information(S102). The final depth information of the reference viewpoint image is determined(S103). The final depth information of a viewpoint image except a reference viewpoint among multi-view points is determined by using the final depth information of the reference viewpoint image (S104). The final depth information of each multi-view image is recorded as a digital image(S105).