An Image selection method from image sequence collected by remotely operated robot for efficient 3D reconstruction

Hanari, Toshihide; Kawabata, Kuniaki; Nakamura, Keita*; Naruse, Keitaro*

Proceedings of International Workshop on Nonlinear Circuits, Communications and Signal Processing (NCSP 2020) (Internet), p.242 - 245, 2020/02

In this paper, we propose an image selection method to realize efficient three-dimensional (3D) reconstruction from image sequence collected by a remotely operated robot. The proposed method is to select adequate images for reconstruction computation based on optical flow measures. We confirmed that the proposed method can select adequate images for 3D reconstruction from the image sequence recorded by the remotely operated robot that is commanded stop-and-go motions. As a result, we demonstrated a 3D reconstruction could be performed efficiently with relatively few images that were selected by the proposed method.