Key points are not available for this paper at this time.
This document highlights on a knowledge-based method to reconstruct planar surfaces of a real indoor environment with occlusion and clutter, which produce a reliable 3D modeling rapidly. The input is 3D point cloud data obtained from a laser scanner, which is known for its accuracy and speed in producing 3D data. The laser is attached on a mobile platform with a servo motor to give a complete 180° coverage, both vertically and horizontally. It takes about 1.7 seconds for the algorithm, which is based on a computational geometry approach, to process and produce the 3D modeling of all existing surfaces on a normal working computer. The method has been tested in different cluttered environments and has shown promising results. This method can be applied towards developing an as-built building Information modeling (BIM) in architectural as well as semantic mapping for a mobile robot Other applications in facility management, building conservation, virtual tours and city infrastructure management could also benefit from this approach.
Shukor et al. (Fri,) studied this question.