Solar Potential Analysis Method Using Terrestrial Laser Scanning Point Clouds

Pengdi Huang, Ming Cheng, Yiping Chen, Dawei Zai, Cheng Wang, Jonathan Li

Abstract: Analysis of sunlight or solar potential requires the data for a targeted scene to be authentic and accessible as much as possible. In this regard, the data for a real-world scene acquired by ground-based laser scanners are comprehensive and convenient, thus potentially meeting this requirement. To get an accurate result and offer an automatic solution for sunlight analysis, this paper proposes a solar potential analysis method that is run directly on 3-D ground laser scanning point clouds. Our method simulates natural illumination, sunlight hours, and solar radiation of the targeted scene for a specified period. This method first extracts the region of interest (ROI) to obtain targeted points. Then, we compute solar position as a virtual light source and propose a control method acting on the ROI to constraint deviation from the point light source. Finally, we adopt the generalized hidden point removal algorithm to cast shadow of obstruction on the ROI. Besides, experiments to validate the shading method results are carried out for three different periods. The quantitative results in the Xiamen case evaluated by the Hausdorff distance demonstrate the advantage and feasibility of our proposed method.