Pose estimation for planar objects in noisy images with an application of UAV automatic landing
หัวหน้าโครงการ
:
มิติ รุจานุรักษ์
ทีมวิจัย
:
มิติ รุจานุรักษ์
หัวหน้าโครงการ
วันที่เริ่มโครงการ
:
1 ก.ค. 2558
วัตถุประสงค์
:
First, as we will use pose estimation to assist landing, our constraint is different from the above mentioned papers. In UAVs, most IMUs (return R) are considered reliable. Hence, what the landing assist system is lacking is the precise T; given IMUs R. GPS is out of question as mentioned earlier. In this regard, we propose to develop a system based on know-how in [Li et al.], [Lu], and information from IMUs (R.)
Second, we would like to target the same problem mentioned in [Schweighofer&Pinz], the real minimal R/T. [Schweighofer&Pinz] already did a geometrical approach to find the 2nd minimum/pose. However, with high level of image noises, the optimal pose by [Schweighofer&Pinz] is still wrong. In this regard, we propose to develop an algorithm fusion method based on know-how from [Schweighofer&Pinz] and Homography.