Xie, YiranLiu, NianjunLiu, ShengBarnes, Nick2015-12-08December 69780769545882http://hdl.handle.net/1885/31886This paper provides a novel and efficient approach to dense stereo matching. The first contribution is, rather than applying disparity distribution inside a segment as hard constraint to directly project the likelihood of corresponding candidates to eachKeywords: Graph cut; Hard constraints; Higher order; Soft constraint; stereo matching; Constraint theory; Graphic methods higher-order; stereo matching201110.1109/DICTA.2011.932016-02-24