The Scan View Planning Algorithm Based on the CAD Model

The surface of large-scale structure is large and complicated, and a lot of different positions under different pose are needed for the measuring system to get integrated three-dimensional surface data. The measurement accuracy could be improved by increasing sampling point number, but there are unavoidable problems like having larger data information and less efficient. When high measurement accuracy and high measurement efficiency are required, it is needed to plan the view point. In this paper, rectangle self-adaption sampling algorithm based on the CAD model is proposed. In this method, firstly, the least sampling points are determined by dividing up the large-scale structure surface area based on the camera measurement area. Secondly, the surface curvature of one view point is calculated and compared with camera incident angle. If all of the surface curvatures of the view point are less than the camera incident angle, the view point is saved. Otherwise, one view point is increased. Finally, the same operations are applied to all of the view points, and the final view point number is determined. Simulation and experimental results show that, the algorithm proposed can measure the surface of large-scale structure quickly and efficiently.