Graph-Cuts Fusion of Distance Fidelity Mapsfor Volumetric Multi-view Stereo

This paper presents a new graph-cuts based volumetricreconstruction technique for multi-view stereo. To refrain from theballoon constraint term widely used in 3D reconstruction approaches, wepropose the using of a more sophisticated and high performance map,namely the distance fidelity map, for graph-cuts optimization. Theapproach first seeks to compute the optimal Distance fidelity map (DFM)of each view using photo-consistency, feature constraint, silhouetteconstraint and smoothness constraint. Occlusion robust and featurerobust mechanism are also considered during this process. After this,according to our proposed objective function, multiple DFMs are fusedinto a graph-cuts framework to give a watertight surface that bestcapture the surface detail. Results demonstrate the high performancereconstruction quality using only 10 views and its generality toextensive actor structures.