石质文物作为文明的重要物质载体，其表面病害检测是文化遗产预防性保护的核心任务之一。然而，面对表面纹理复杂、风化程度不一的点云数据，现有检测分割算法存在两方面不足：1）传统硬阈值方法依赖人工调参、适应性差；2）人工纹理与自然病害在几何特征上高度相似，导致分割过程中两者易粘连混淆。为此，本文提出一种融合统计模型与几何拓扑分析的自动化检测方法。该方法首先选取几何粗糙度与颜色一致性构建二维特征向量，通过高斯混合模型实现无监督自适应异常提取，替代传统依赖人工调参的硬阈值方法；随后提出多约束的局部凸包连接片（locally convex connected patches,LCCP）分割算法，通过引入正交距离约束与法向夹角约束，从几何拓扑层面解决纹理与病害的分割粘连问题，并进行后处理优化分割结果。在3处典型病害与纹理粘连测试区上验证可知，本文方法相较于传统LCCP算法与区域生长算法，精确率、F1值和交并比等指标均全面提升，3个测试区F1值分别达到0.915 8、0.906 3和0.926 0，有效解决纹理与病害的粘连难题，为石刻文物病害数字化检测提供了可靠方案。
Stone cultural relics are important material carriers of civilization, and detecting their surface damage is a core task in the preventive conservation of cultural heritage.However, when dealing with point cloud data characterized by complex surface textures and varying degrees of weathering, existing detection and segmentation algorithms exhibit two limitations: 1) traditional hard-threshold methods rely on manual parameter tuning and exhibit poor adaptability; 2) artificial textures and natural weathering damage are highly similar in geometric features, which leads to region merging and confusion between them during segmentation. Therefore, an automatic detection method combining a statistical model and geometric topological analysis was proposed. First, geometric roughness and color consistency were selected to construct a two-dimensional feature vector, and unsupervised adaptive anomaly extraction was realized by a Gaussian mixture model to replace the traditional hard-threshold method relying on manual parameter tuning. Subsequently, a multi-constraint locally convex connected patches(LCCP) segmentation algorithm was designed. By introducing an orthogonal distance constraint and a normal angle constraint, it solved the region-merging problem between textures and damage from a geometric-topological perspective, and a post-processing step was conducted to optimize the segmentation results. Validation was carried out on three typical test regions with texture–damage merging. Compared with the traditional LCCP algorithm and region growing algorithm, indicators of the proposed method such as precision, F1-score, and intersection over union were comprehensively improved; the F1-scores on the three test regions reached 0.915 8, 0.906 3, and 0.926 0, respectively. The proposed method effectively resolved the merging problem between textures and damage, and provided a reliable solution for the digital detection of damage in stone cultural relics.
[1] 陈晓瑜, 张婷, 王金华. 我国石质文物分类研究[J]. 石窟与土遗址保护研究, 2022, 1(1): 39-46. Chen X Y, Zhang T, Wang J H. Study on the classification of stone cultural relics in China [J]. Research on the Conservation of Cave Temples and Earthen Sites, 2022, 1(1): 39-46. (in Chinese)
[2] 孙弘仪. 露天石质文物的本体保护及外界环境研究[J]. 文物鉴定与鉴赏, 2025(19): 38-41. Sun H Y. Study on the ontology protection and external environment of open-air stone cultural relics [J]. Identification and Appreciation to Cultural Relics, 2025(19): 38-41. (in Chinese)
[3] 刘远坚, 韩绍康, 张程, 等. 潮湿环境砖石文物本体病害分析及保护技术研究进展[J]. 重庆大学学报(自然科学版), 2024, 47(10): 120-137. Liu Y J, Han S K, Zhang C, et al. A comprehensive review of the disease analysis and conservation techniques for brick and stone artifacts in humid environments [J]. Journal of Chongqing University (Natural Science Edition), 2024, 47(10): 120-137. (in Chinese)
[4] Zou X X, Sneed L H, Mirala A, et al. Debonding detection of defected CFRP-concrete interface using active microwave thermography [J]. Composite Structures, 2023, 310: 116753.
[5] Jung J, Kim T, Min H, et al. Intricacies of opening geometry detection in terrestrial laser scanning: an analysis using point cloud data from BLK360[J]. Remote Sensing, 2024, 16(5): 759.
[6] 肖开荣, 孟庆祥, 杨鹤, 等. 一种基于LiDAR多视轮廓点的三维壁画变化检测方法[J]. 应用科学学报, 2024, 42(5): 733-746. Xiao K R, Meng Q X, Yang H, et al. Change detection of 3D mural surface based on multiview contour points of LiDAR data [J]. Journal of Applied Sciences, 2024, 42(5): 733-746. (in Chinese)
[7] 闫振国, 陈杨, 刘如飞, 等. 基于空间临近的森林样地树木点云分割方法[J]. 应用科学学报, 2025, 43(6): 1003-1014. Yan Z G, Chen Y, Liu R F, et al. Point cloud segmentation method for trees in forest sample plots based on spatial proximity [J]. Journal of Applied Sciences, 2025, 43(6): 1003-1014. (in Chinese)
[8] Jiang H B, Li Q, Jiao Q S, et al. Extraction of wall cracks on earthquake-damaged buildings based on TLS point clouds [J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2018, 11(9): 3088-3096.
[9] Wood R L, Mohammadi M E. Feature-based point cloud-based assessment of heritage structures for nondestructive and noncontact surface damage detection [J]. Heritage, 2021, 4(2): 775-793.
[10] Cao T, Hu J Y, Liu S. Enhanced edge detection for 3D crack segmentation and depth measurement with laser data [J]. International Journal of Pattern Recognition and Artificial Intelligence, 2022, 36(5): 2255006.
[11] Gu J, Zhang Y W, Yin Y X, et al. Surface defect detection of cabbage based on curvature features of 3D point cloud [J]. Frontiers in Plant Science, 2022, 13: 942040.
[12] Stałowska P, Suchocki C, Rutkowska M. Crack detection in building walls based on geometric and radiometric point cloud information [J]. Automation in Construction, 2022, 134: 104065.
[13] Deng L, Sun T, Yang L, et al. Binocular video-based 3D reconstruction and length quantification of cracks in concrete structures [J]. Automation in Construction, 2023, 148: 104743.
[14] Shibano K, Morozova N, Ito Y, et al. Evaluation of surface damage for in-service deteriorated agricultural concrete headworks using 3D point clouds by laser scanning method [J]. Paddy and Water Environment, 2024, 22(2): 257-269.
[15] Dong Q, Wang S K, Chen X Q, et al. Pavement crack detection based on point cloud data and data fusion [J]. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 2023, 381(2254): 20220165.
[16] Gamal A, Wibisono A, Wicaksono S B, et al. Automatic LiDAR building segmentation based on DGCNN and euclidean clustering [J]. Journal of Big Data, 2020, 7: 102.
[17] Chen J D, Cho Y K. CrackEmbed: Point feature embedding for crack segmentation from disaster site point clouds with anomaly detection [J]. Advanced Engineering Informatics, 2022, 52: 101550.
[18] Hu K W, Chen Z, Kang H W, et al. 3D vision technologies for a self-developed structural external crack damage recognition robot [J]. Automation in Construction, 2024, 159: 105262.
[19] Feng H F, Li W, Luo Z P, et al. GCN-based pavement crack detection using mobile LiDAR point clouds [J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(8): 11052-11061.
[20] Stein S C, Schoeler M, Papon J, et al. Object partitioning using local convexity [C] //IEEE Conference on Computer Vision and Pattern Recognition, 2014: 304-311.
