A Novel Classification & Repairing Framework For Irregular Model Reconstruction Using Mass Point Clouds Based On Fuzzy Inference

Mass point clouds of irregular model obtained from laser scanner are incomplete, which causes the problem that affects surface reconstruction and subsequent application. Due to the fact that geometry feature of irregular model local area has uncertainty, it is difficult to adopt one effective method to repair the lost data of model. In this paper, point clouds of irregular model are meshed firstly. According to the shape of the boundary, the hole is divided into different types by fuzzy inference system (FIS). Then the corresponding repairing method is presented for each type. For these repairing methods, public modules are introduced in detail. The repairing effect is improved greatly for corresponding methods are carried out to repair hole with different shapes. Experimental results show that the efficiency and accuracy of dental models are improved by the proposed classification and repairing framework effectively.