EVALUATION OF THE QUALITY OF REAL-TIME MAPPING WITH CRANE CAMERAS AND VISUAL SLAM ALGORITHMS

Abstract. In the context of development an autonomous tower crane, usage crane cameras to map respective workspace as basis for path planning is investigated. The goal generate up-to-date DEM input control. As construction sites are highly dynamic scenes, it crucial be able react any changes. Thus, real-time mapping with a visual SLAM solution aspired. quality important such safety critical application, we evaluating four state-of-the-art solutions, namely ORB-SLAM3, LDSO, DSM and DROID-SLAM. results show that all approaches can handle our specific camera setup thus generally suited application. accuracies tested methods even competitive result from standard offline photogrammetric processing, at least major part test site. However, there limitations regards completeness. Consequently, investigations deliver good accurate mapping, but their application further refinements have made.