Position and velocity measurement method from a single image using modulated illumination
Kentaro Matsuo, Takahiro Yakoh · 2018
Visual feedback systems utilize position and velocity information estimated from frames captured by a camera. Since the velocity information is derrived through pseudo differential with two consecutive frames, the information must be delayed essentially, and the delay often limits the performance of visual feedback systems. To eliminate the delay, this paper proposed a novel technique to estimate position and velocity information from a single image. By detecting the start and end points of the motion blur of a target object in an image, the velocity information during the exposure time of the image can be estimated fundamentally. To tag such a temporal information on motion blurs, a modulated illumination was employed. This tagging method aimed to make motion blur detection robust against various background scenery. Some simulation results validated the feasibility of the proposed method.