Point cloud generation of a building from close range thermal images

Abstract

Thermal inspection of a building is a very important part of facility management. This paper presents the methodology how to capture thermal images on a case of a family house, how to process thermal images using current photogrammetric method and how to achieve point cloud where each point is enhanced by information about temperature. This enhanced point cloud could be data for BIM or GIS and could be a valuable result for facility managers where they can detect thermal leaks or anomalies. Due to the point cloud, the thermal leaks or anomalies could be easily located and described. The accuracy of the point cloud is tested by check points and the resulted point cloud is compared with point cloud gathered by a laser scanner. The last part of the paper is devoted to a discussion about the results and outputs of the photogrammetric processing of thermal images.