3D Digitisation of Large-Scale Unstructured Great Wall Heritage Sites by a Small Unmanned Helicopter

摘要

The ancient Great Wall of China has long suffered from damage due to natural factors and human activities. A small low-cost unmanned helicopter system with a laser scanner and a digital camera is developed to efficiently visualize the status of the huge Great Wall area. The goal of the system is to achieve 3D digitisation of the large-scale Great Wall using a combination of fly-hover-scan and flying-scan modes. However, pose uncertainties of the unmanned helicopter could cause mismatching among point clouds acquired by each hovering-scan. This problem would become more severe as the target area becomes larger and more unstructured. Therefore, a hierarchical optimization framework is proposed in this paper to achieve 3D digitisation of the large-scale unstructured Great Wall with unpredictable pose uncertainties of the unmanned helicopter. In this framework, different optimization methodologies are proposed for the fly-hover-scan and flying-scan modes, respectively, because different scan modes would result in different features of point clouds. Moreover, a user-friendly interface based on WebGL has been developed for 3D model visualization and comparison. Experimental results demonstrate the feasibility of the proposed framework for 3D digitisation of the Great Wall segments.