Multi-scale large component assembly gap point cloud filtering method

The invention relates to a multi-scale large component assembly gap point cloud filtering method, which comprises the following steps of: performing ROI region segmentation on an original point cloud P of a large component assembly gap, and extracting to obtain an ROI region point cloud # imgabs1 # with the minimum point cloud density # imgabs0 # through a point cloud density sensor so as to retain effective point cloud data; performing invalid point filtering and large-scale outlier filtering on the point cloud # imgabs2 # of the ROI region to obtain a filtered point cloud # imgabs3 #; clusters of all point clouds of # imgabs4 # are extracted through a region growing clustering method, so that a point cloud cluster with continuity is formed, gap-independent point clouds are segmented and filtered based on a weight voting method, effective point clouds # imgabs5 # are obtained, and the accuracy of the gap-related point clouds is improved; and performing small-scale point cloud fairing on the effective point c