Multi-view stereo depth-map merging for the Three-Dimensional Swept Volume Display

This paper presents a multi-view stereo depth-map merging algorithm for the Three-Dimensional (3D) Swept Volume Display. In this paper, we first transform the disparity map into the depth-map and do the two-dimensional filtering based on the continuity of the object´s surface, then merging each filtered depth-map in the form of the first camera coordinate system, Finally, the repetitive voxel is filtered by 3D volumetric filter. Experimental results show that we can get a dense 3D model which could be displayed in the 3D Swept Volume Display with full 360 degree view angles.