Estimating Camera Position and Posture by Using Feature Landmark Database

Abstract

Abstract. Estimating camera position and posture can be applied to the fields of augmented reality and robot navigation. In these fields, to obtain absolute position and posture of the camera, sensor-based methods using GPS and magnetic sensors and vision-based methods using input images from the camera have been investigated. However, sensor-based methods are difficult to synchronize the camera and sensors accurately, and usable environments are limited according to selection of sensors. On the other hand, vision-based methods need to allocate many artificial markers otherwise an estimation error will accumulate. Thus, it is difficult to use such methods in large and natural environments. This paper proposes a vision-based camera position and posture estimation method for large environments, which does not require sensors and artificial markers by detecting natural feature points from image sequences taken beforehand and using them as landmarks.

Augmented Reality Aided Operation and Maintenance System for Indoor Environments

“…1 (a)) to detect the position and pose of the AR content, which can ruin the scenery or a manual document that contains the AR content. AR systems without artificial markers have also been proposed (e.g., [4][5][6]). Systems without markers preserve the scenery but the accuracy is lower than an AR system with artificial markers.…”

Wide-area scene mapping for mobile visual tracking

“…Overall, our system is more flexible because it does not require an omnidirectional camera on the tracked mobile device. Oe, Sato and Yakoza use a reconstruction from omnidirectional video to track a video feed, but do not consider a real-time approach to the system [20].…”

Balancing Design Freedom and Constraints in Wall Posters Masquerading as AR Tracking Markers

“…On the other hand, registration methods using natural feature points in a real scene [8][9][10][11] differ from the fiducial marker methods. These methods do not affect the environment, and on comparing with the fiducial marker methods, most of them do not show stability and robustness.…”

“…These methods do not affect the environment, and on comparing with the fiducial marker methods, most of them do not show stability and robustness. These methods can be classified roughly into the following two technologies: 1) camera tracking without any environmental information [8,9], and 2) measuring the absolute position and angle of the camera using a pre-constructed database of environmental information [10,11]. Simply putting, the former technology estimates the camera position of the current frame, relative to that of the previous frame, by comparing the current image and previous camera images.…”