Automated Detection of Road Manhole Covers from Mobile LiDAR Point-Clouds Based on a Marked Point Process

This paper presents a novel algorithm for detecting road manhole covers from mobile LiDAR point-clouds. This algorithm takes advantage of a marked point process of discs and rectangles to model the locations and geometric structures of the manhole and sewer well covers. The algorithm also uses the Bayesian paradigm to obtain a posterior distribution for the marked point process conditional on the geo-referenced intensity image. A Reversible Jump Markov Chain Monte Carlo (RJMCMC) algorithm is implemented to simulate the posterior distribution. Finally, the maximum a posteriori (MAP) scheme is used to obtain an optimal detection. This algorithm has been examined by a set of mobile LiDAR point-clouds acquired by a RIEGL VMX-450 mobile laser scanning system. The results demonstrate the efficiency and feasibility of the proposed algorithm for automatically detecting road manhole and sewer well covers.