Image based modeling via plane sweep based surface growing

In this paper, we present a multi-view stereo based shaped modeling method. Using images captured from different viewpoints, our approach can provide objects' 3d models with high fidelity details automatically and efficiently. We firstly use a strict plane based sweep stereo method via GPU to compute quasi-dense depth maps which usually have many holes. Then, a simplified patch based surface growing method is used to compute dense depth maps and the corresponding 3d geometry model. Different from other multi-view stereo methods, we do not optimize object normal during expansion process, but gradually compute the normal information from reconstructed quasi dense neighbors. This makes our method works well on much more difficult scenarios, such as textureless, wide baseline, varying illumination than plane based sweep methods and more efficient than surface growing methods. Experiments show that our method can generate high fidelity 3D object shape model quite efficiently.