Detail-feature-preserving surface reconstruction

摘要: In this paper, we propose a feature-preserving surface reconstruction method from sparse noisy 3D measurements such as range scanning or passive multiview stereo. contrast to earlier methods, define novel type of explicit filter—regularized weighted least squares filter—to characterize the detail features wrinkles and sharp features. To account for noise, rasterize input-oriented points into probabilistic volume (base volume) then create guidance by Gaussian filtering. Both base are further filtered regularized filter detect recover After two-stage filtering, global minimal is computed graph cut meshed geometric model. Experimental results on various datasets show that our robust outliers, missing parts, which makes it more suitable fit indoor/outdoor stereo data. Unlike other can completely scene structures preserve point samples. Copyright © 2012 John Wiley & Sons, Ltd.