Volume 3,Issue 7
基于法向量和曲率估计的点云数据预处理研究
针对复杂点云在三维扫描与模型采样过程中存在的噪声干扰、点密度不均、局部几何属性缺失等问题，提出一种以法向量估计和曲率计算为核心，并结合统计滤波去噪、体素网格下采样的点云预处理流程。研究表明，该预处理流程能够改善点云输入质量，提高局部几何特征的一致性，为后续复杂零件点云深度学习特征提取、鲁棒匹配和高精度配准提供稳定的数据基础。
[1] ZHANG Y X, GUI J, YU B, CONG X F, GONG X, TAO W B, TAO D. A comprehensive survey and taxonomy on point cloud registration based on deep learning[C]//Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence. Jeju: IJCAI, 2024: 8344-8352.
[2] CHEN L, FENG C Z, MA Y P, ZHAO Y K, WANG C R. A review of rigid point cloud registration based on deep learning[J]. Frontiers in Neurorobotics, 2024, 17: 1281332.
[3] HUANG S Y, GOJCIC Z, USVYATSOV M, WIESER A, SCHINDLER K. PREDATOR: registration of 3D point clouds with low overlap[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Nashville: IEEE, 2021: 4267-4276.
[4] YEW Z J, LEE G H. REGTR: end-to-end point cloud correspondences with transformers[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans: IEEE, 2022: 6667-6676.
[5] LI D W, WEI Y C, ZHU R S. A comparative study on point cloud down-sampling strategies for deep learning-based crop organ segmentation[J]. Plant Methods, 2023, 19: 124.
[6] LANG I, MANOR A, AVIDAN S. SampleNet: differentiable point cloud sampling[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 7578-7588.
[7] LUO S T, HU W. Score-based point cloud denoising[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. Montreal: IEEE, 2021: 4583-4592.
[8] MAO A H, DU Z H, WEN Y H, XUAN J, LIU Y J. PD-Flow: a point cloud denoising framework with normalizing flows[C]//European Conference on Computer Vision. Cham: Springer, 2022.
[9] ZHU R S, LIU Y, DONG Z, JIANG T P, WANG Y, WANG W P, YANG B S. AdaFit: rethinking learning-based normal estimation on point clouds[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. Montreal: IEEE, 2021: 6118-6127.
[10] XIU H Y, LIU X, WANG W M, KIM K S, MATSUOKA M. MSECNet: accurate and robust normal estimation for 3D point clouds by multi-scale edge conditioning[C]//Proceedings of the 31st ACM International Conference on Multimedia. New York: ACM, 2023: 2535-2543.