We study the problem of feature extraction in point cloud registration. Traditional point clouds has the characteristic of irregular structure, which causes the neighborhood relationship that cannot effectively obtain point cloud data, and increases the difficulty of feature extraction in the point cloud registration task. This paper proposes a graph convolution point cloud registration network based on a deformable kernel. Compared with the non-deformable kernel, the proposed network is more suitable for irregular and unstructured point cloud data. Meanwhile, the network uses the semantic residual module to restore the lost local information and enhance the integrity of feature expression. The feature fusion layer integrates global and local features to enhance the model’s ability to express the features of complex point cloud data. We conducted tests on the 3DMatch, 3DLoMatch, and KITTI datasets to verify the effectiveness of the algorithm.
Niu et al. (Fri,) studied this question.