View Planning Method for Automatic Measurement and Reconstruction of Unknown Objects

In order to the automatic measurement and reconstruction of the unknown three dimensional objects, a new view planning method based on the limit visible region of the vision system is proposed in this paper. Firstly, the visible and measured region of the monocular line-laser vision system is obtained, and the limit visible surface is modeled by the known boundary data obtained from initial view and is used to predict the surface information of the unknown object. Then the visibility criterion of next viewpoint is determined according to the predict curve, and the position that could obtain the maximum visible curve length in left or right planning process is defined as the next best view(NBV) candidate position. The larger one is selected as the final NBV position by comparing the visible area of the above two candidate positions. Finally, we have experimentally validated our research with some experiments for reconstruction.