Loading...
Please wait, while we are loading the content...
Similar Documents
Real-time 3D Pose Estimation and Tracking 3D Marker using Dual-eyes Camera for Underwater Vehicle
| Content Provider | Semantic Scholar |
|---|---|
| Author | Myint, Myo |
| Copyright Year | 2015 |
| Abstract | Recently a number of research have been conducted on pose estimation and tracking object using visual information 1)-5). Each of them is with different merits and certain considerations. Most of them are based on the idea of image reconstruction using monocular vision. Normally some features are extracted and matched with the corresponding ones in other images to get the information of the object such as pose, size, position and so on. But they are non feasible in term of computational load as well as reliability of mapping features. Therefore, according to challenges in terms of computational burden, real-time pose estimation and tracking object poses the difficulties to researchers. In this paper, model-based recognition approach is applied because of its performance in terms of less sensitive on camera calibration, comparing to other methods like feature based recognition in which pose of the target object should be determined by a set of image points, complex searching the corresponding points and time consuming. We have developed a visual servoing system that recognizes the relative position and orientation between ROV and target object utilizing GA using images from dual-eyes camera mounted in the ROV. We conducted regulated performance experiments on the relative position and orientation control between the ROV and 3D target. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.suri.sys.okayama-u.ac.jp/underwater/pdf/thesis/pdf2017/tracking%20paper%20%5B1%5D%20for%20video%207%20.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
