1000-fps target tracking using vibration-based image features

In this paper, we propose a concept for the real-time target tracking of vibrating objects to extract vibration image regions using digital filters at all the pixels in an image, assuming that periodic changes in image intensities exist at pixels around vibrating objects. By combining vibration-based processing with color extraction as a masking process, our algorithm can prevent the latency effect on a digital filter, which may degrade its tracking performance. Digital filters at a pixel level are implemented on a 1000-fps vision system that can measure vibration distribution over a wide range of frequencies. On our target tracking system, we can track a vibrating object at the center of the camera view by controlling a 2-DOF (Degree Of Freedom) active vision system with vibration-based image features. We have verified its effectiveness by presenting several experimental results using objects vibrating at high frequencies, which cannot be observed with the naked eye.