MOTION DETECTION and TRACKING in CONSTRUCTION SPACE USING TEMPORAL POINT CLOUDS

Conventional ICT construction vehicles have been developed to improve productivity. However, the conventional ICT construction vehicles have several technical issues in cooperated operations with workers. Therefore, we focus on a methodology to keep the safety in cooperated operations with workers and construction vehicles at a construction site. Our proposed methodology is based on 3D sensing from a construction. In 3D sensing, temporal point clouds are acquired for moving object tracking. Moreover, we develop an exterior-typed 3D measurement unit with low-price RTK-GNSS and LiDAR to provide lower-price ICT construction vehicles. In this research, we experimented excavation works using a backhoe at a simulated construction space. Through the experiment, we confirmed that our methodology can detect and track the bucket and workers using temporal point clouds acquired from a backhoe in drilling works.