主催: 一般社団法人 日本機械学会
会議名: ロボティクス・メカトロニクス 講演会2018
開催日: 2018/06/02 - 2018/06/05
This study proposes a novel concept of undersampled-video-based vibration source localization that can largely reduce the computational complexity in pixel-level frequency filtering. We discuss how to determine the frame rate of the undersampled video sequence in which the target objects vibrating at different frequencies can be distinguished as well as DC background components when the vibration frequencies are much larger than the Nyquist frame frequency of the video sequence. The performance of undersampled-video-based vibration source localization is demonstrated by several offline experimental results for multicopters with their propellers rotating at hundreds of Hertz.