光学精密工程2025，Vol.33Issue(20)：3315-3330,16.DOI:10.37188/OPE.20253320.3315
自适应采样与几何-空间特征融合的点云配准
Point cloud registration using adaptive sampling and geometric-spatial feature fusion
摘要
Abstract
Point cloud registration in 3D reconstruction scenarios faces significant challenges,as traditional local feature descriptors often fail due to insufficient keypoints,weak geometric descriptiveness,and poor matching robustness.To address these issues,this study proposed an adaptive sampling and geometry-spatial feature fusion algorithm.First,adaptive density-based voxelization followed by nearest-neighbor downsampling was proposed to address size and density imbalances between low-overlap point cloud pairs while achieving efficient data reduction.Next,surface normals were computed via KD-tree search,and a filtering mechanism incorporating neighborhood point count and linearity constraints was employed to iden-tify salient keypoints.These selected points were subsequently encoded using fused geometry-spatial de-scriptors to overcome the redundancy and weak descriptiveness of conventional methods.Finally,a bidi-rectional correspondence approach based on histogram similarity identified reliable point matches,which were then refined through RANSAC to attain robust,high-precision registration under low-overlap condi-tions.The algorithm was validated on public benchmarks and real-world datasets.Experimental results demonstrate that our method reduces average matching error by 51.14%,64.16%,and 78%compared to ISS+3DSC+K4PCS,ISS+FPFH+RANSAC,and TOLDI+RANSAC,respectively.Additional-ly,our approach achieves the highest runtime efficiency among all compared methods,evidencing superior accuracy,adaptability,and robustness.关键词
点云配准/低重叠率/特征融合/三维重建Key words
point cloud registration/low overlap/feature fusion/3D reconstruction分类
信息技术与安全科学引用本文复制引用
张伟,方麒,曾志龙,桂冠,宋杰,廉文波,胡小亮,王生怀,王宸..自适应采样与几何-空间特征融合的点云配准[J].光学精密工程,2025,33(20):3315-3330,16.基金项目
国家自然科学基金资助项目(No.52405590,No.52475557,No.51475150) （No.52405590,No.52475557,No.51475150）
国家科技重大专项(No.2018ZX04027001) （No.2018ZX04027001）
