Dealing With Massive Volumetric Visualization

摘要

Shape from shading is a method for recovering the shape of a surface using the brightness information of a single image. This paper presents a new approach to three mesh surface reconstruction using Radical basis function neural networks(RBFNN).It includes three steps: first, the cloud data are preprocessed for smoothing; second, feature lines are extracted and the cloud data are segmented; at last, NURBS surface patches are created over rectangular mesh and trimmed to form an entire surface using RBFNN.The proposed approach is based on using the neural networks as a parametric representation of the three-dimensional object and the shape from shading problem is formulated as the minimization of an intensity error function with respect to the network weights. Moreover, we demonstrate the robustness of our approach to strong illumination variations and with significant pose variations. The recovered face shape is then shown along with the original surface.