Path planning for surface inspection on a robot-based scanning system

This paper proposes an automatic model-based path planning method for surfaces inspection on a robot-based scanning system. The original CAD model of the surface is a triangular mesh model. The scanning system consists of a 6 degree of freedom robot and a coded structured light scanner. Based on the system structure and the working conditions of the scanner, some scan constraints, i.e. field of view, working distance, view angle, and overlap, are concerned in path planning. The proposed method consists of four steps: model framework extraction, scan region segmentation, viewpoint generation, and scan path generation. Firstly, model framework is extracted from the projection of model in main direction. Secondly, the model is divided into several scan regions based on the framework. Thirdly, viewpoints satisfied scan constraints are planned for each scan region. Lastly, a scanning path is generated passing through these viewpoints. Experiments are presented to demonstrate the feasibility and advantages of the proposed method.