Creating Three-Dimensional Thermal Maps

摘要

A method for generating 3D maps of mines with thermal imaging texture from a robotic platform is presented. The objective is to use these models to generate risk maps that can be used to assess mine safety. Analysis will take place offline, but the data collection platform will be autonomous. Our registration method is based on aligning 3D descrip- tors that are extracted from range images and uses ICP for refinement. The descriptors encode the distribution of radial distances in the vicinity of keypoints, and enable alignment of non-sequential scans that are visibly different. Thermal texture is added to the registered point cloud using a once-off calibration between the 3D and thermal cameras. Two alternate visualisation schemes are discussed, and results are shown for a real mine stope. I. INTRODUCTION The CSIR is pursuing the development of an Underground Autonomous Mine Safety Platform (UAMSP) (1) that will provide insight into aspects of mine safety. In particular, the vehicle will collect thermal and electronic sounding data that can be used to assess hanging wall stability (2), thereby indicating the risk of potential rock falls — a significant cause of injury and death in mines. One of the challenges involves building a 3D map of the mine stope and overlaying thermal measurements in order to generate a risk map. A framework for registering 3D and thermal data in order to accomplish this task is presented. A. Background Previous studies in South African hard rock mines (3), (4), (5) have identified a correlation between loose rock and its surface temperature gradient. Blasting often results in the for- mation of cracks that separate portions of rock which interrupts the heat flow and causes preferential cooling in a localised area. The effect can be detected by long-wave infrared cameras (thermal imaging) and used to identify candidate areas that are unsafe. Electronic sounding, which requires contact with the surface, can then target these candidate areas to provide more detailed analyses. The fact that thermal differences can be caused by a number of phenomena, such as surface angle and material composition, motivated a multi-sensor approach. The incorporation of both 3D and thermal imaging cameras enables thermal risk maps to correctly account for surface geometry and provides a mecha- nism for mapping the environment. (An important component for assisting robot navigation and facilitating visualisation.)