Host: The Institute of Image Electronics Engineers of Japan
Co-host: The Institute of Image Information and Television Engineers, The Society for Art & Science, The Computer Graphic Arts Society(CG-ARTS)
Name : Reports of the 285th Technical Conference of the Institute of Image Electronics Engineers of Japan
Number : 285
Location : [in Japanese]
Date : March 16, 2018 -
Image feature keypoints of a natural image which is used as an AR marker are not uniformly. Furthermore, it is challenging problem to detect the marker which is changed such as folding or bending. Previous methods related with this study are proposed. For example, they are a method for detecting a nonplanar AR marker using depth information and a method of tracking the shape deformation of the marker from the state of the flat surface using movie. This paper aims to detect the nonplanar marker with a single camera. In order to estimate the shape of the nonplanar marker at the same time of the detection, we use keypoints of the AR marker divided into grid mesh. It, however, is difficult to detect the marker with the meshes, since there are not many the keypoints on each mesh. Assuming the keypoints of the marker are not uniformly, this paper proposes nonplanar marker detection with adaptive division mesh considering distribution of the keypoints and a method to complement the undetected meshes from the detected meshes.