Host: The Japan Society of Mechanical Engineers
Name : [in Japanese]
Date : June 02, 2018 - June 05, 2018
This paper proposes an intelligent and fast tracking method for robust trackability against appearance changes that hybridizes a correlation-based tracking algorithm operating at hundreds of fps with a deep-learning-based recognition algorithm at dozens of fps. A prototype of intelligent mechanical tracking system was developed by implementing our hybridized tracking algorithm on a 500-fps vision platform, and a complex-shape target can be robustly tracked at the center of the camera view in real time by controlling a pan-tilt active vision system with 500 Hz visual feedback. Its performance was verified by showing experimental results for a pre-learned object, which was quickly manipulated by human hand in complicated background.
