Slum Terrain Mapping Using Low-Cost 2D Laser Scanners

Abstract

This paper presents a motorbike-based custom made scanning and mapping system for surveying slums and highly populated urban regions. These vicinities are difficult to reach through standard vehicular scanning systems and require a compact solution as presented in this paper. The system consists of two small range 2D Hokuyo laser scanners mounted in right angle orientations to capture the environment. In addition, the global positioning system, the wheel encoder, the inertial measurement unit, and cameras have been integrated with the system to estimate the pose and visual information. Sensorial information has been used to localise the system using Kalman Filtering. Later, by applying the standard transformations, the 3D point cloud map of the surveyed vicinity has been developed. The scanning system has been tested at various locations including densely populated and slum regions. Precise and detailed 3D mapping results have been obtained, which are further extensively analysed to understand the built structure and the road furniture. The working of the system is found to be quite economical and faster than that of local urban surveying systems.