A framework development for mapping and detecting changes in repeatedly collected massive point clouds

On a construction site, progress deviations can cause fatal damages to the project and stakeholders. Therefore, accurately monitoring and managing the construction environment is essential. Efficiently detecting and recognizing the changes will be the key factor for monitor and management goals. In this research, we present a framework for detecting and recognizing changes in repeatedly collected, massive 3D point cloud data from a mobile laser scanning system. The framework mainly consists of three parts; 1) mapping system; 2) analysis system; 3) Hadoop platform. Collecting point clouds is repeatedly executed to detect the changes over different time epochs. For collecting point cloud data, a mobile laser scanning system was developed based on Robot Operating System (ROS). Detecting changes between repeatedly collected point clouds have been processed based on Hadoop platform. Finally, detected changes are then implemented to a semantic mapping process which is based on deep learning. Developed framework have potential for wide application in massive point cloud data processing, construction site monitoring, street level change detection, and facility management.