Leak-Mending and Recruitment of Incomplete Points Data In 3D Reconstruction Based on Genetic Algorithm

The leak-mending of mass points data and the recruitment of incomplete points data have great impact on the 3D reconstruction. In this paper, we report an approach to the analysis and reconstruction of 3D scanning points data based on the genetic algorithm to explore the optimum value of recruiting data by controlling the normal vector of the triangle surface, in which the deskpoint could be cleared by boundary analysis of the points data. The leak-mending and recruitment of incomplete points data would be processed by taking the variable of normal vector as the objective function and keeping the smoothness of curved surface. The effectivity of this approach to the points data processing has been confirmed by the results of experimental samples.