High-Speed Orientation Measurement of a Moving Object with Patterned Light Projection and Model Matching

Abstract

Vision is a versatile and powerful sensory function for a robot to understand its surrounding environment and to decide an action to be executed. In recent years, three-dimensional shape measurement of not only stationary objects but also moving objects has been extensively studied. In our laboratory, we developed a high-speed three-dimensional sensor that performs depth measurements at 500 fps. In this research, we aimed to recognize a moving object with this high-speed three-dimensional sensor and use it for motion generation of a robot. The measurement accuracy was evaluated by implementing orientation measurement of a moving object in real time at 500 fps with the model matching method.