Depth map generation from a single image using local depth hypothesis

Abstract

This paper proposes an interactive method of depth map generation from a single image for 2D-to-3D conversion. First, the proposed method groups an input image into similar regions to preserve details and segments the image into salient regions with user interaction. Then, the proposed method generates local depth hypothesis using structural information of the input image and salient regions. Finally, the proposed method generates depth map using local depth hypothesis and grouped regions. Experimental results show that the proposed method gives more natural depth map in terms of human perception. It can be applied to depth image based rendering such as stereo view rendering.