Search

Multiple-view stereo reconstruction is a key step in image-based 3D acquisition and patchmatch based method is suited for large scale scene reconstruction. In this paper we extend the two-view patchmatch stereo to multiple-view in the multiple-view stereo pipeline. The key of the proposed method is to select multiple suitable neighboring images for a reference image, compute the depth-maps and merge the depth-maps. Experimental results on benchmark data sets demonstrate the accuracy and efficiency of the proposed method.

We present a method for 3D shape reconstruction of inextensible deformable surfaces from a single image. The key of our approach is to represent the surface as a 3D triangulated mesh and formulate the reconstruction problem as a sequence of Linear Programming (LP) problems. We use a closed-form method to generate an initial structure, then refine this structure by solving the LP problem iteratively. The robustness and accuracy of our approach are evaluated qualitatively on real data.