OSIRIS-REx OLA point cloud registration based on keypoints matching

Abstract

The OSIRIS-Rex Laser Altimeter (OLA) is the first scanning lidar instrument to fly a planetary mission. The OLA scans Bennu for about a month during the Orbit B mission phase and obtains 911 frames of point clouds. Due to the uncertainty of spacecraft position and pointing, there will be offsets between overlapping point clouds. In our method, the point cloud is first projected onto a plane, and then the keypoints are extracted using the SIFT algorithm. Finally, we perform coarse and global adjustments based on keypoints. However, low accuracy of the corresponding keypoints will lead to bad registration. In order to improve the accuracy of keypoints matching and point cloud registration, we use the tuple test and RANSAC algorithm to eliminate mismatched points. For the overlapping point clouds of two frames, the RMSE between keypoints is about 0.04m after registration. The results show that this method can improve the accuracy of point cloud registration to a certain extent and meet the application requirements.