A Novel Sketch-Based Registration Framework for Point Cloud Frames.

Gang Ma,Hui Wei

DOI: https://doi.org/10.1109/lgrs.2023.3299562

IF: 5.343

2023-01-01

IEEE Geoscience and Remote Sensing Letters

Abstract:Point cloud registration is a crucial part of 3-D computer vision. Existing point cloud registration methods primarily concentrate on utilizing features such as points, lines, and planes, disregarding the valuable contour cues inherent in the scene. In this article, we propose a novel sketch-based framework for point cloud registration that incorporates contour cues to enhance the point cloud registration task. To fully exploit the abundant information provided by contour cues in the scene, the point cloud is first abstracted into a sketch consisting of contour cues obtained through the utilization of planar features, which greatly preserves the inherent contour information. Subsequently, a local contour geometric descriptor is introduced to encode the contour cues in the sketch. Finally, a voting-based contour point pair feature (CPPF) framework is employed to fuse planar features, local contour geometric features, and point pair geometric features, enabling precise estimation of the pose transformation between pairwise point clouds. Extensive experiments conducted on two large-scale outdoor point cloud datasets and two indoor point cloud datasets validate the effectiveness of the proposed sketch-based method. Our proposed method successfully suppresses rotation and translation errors, ultimately achieving state-of-the-art performance.