English

应用科学学报 ›› 2026, Vol. 44 ›› Issue (4): 585-597.doi: 10.3969/j.issn.0255-8297.2026.04.005

• 智能视觉感知 • 上一篇    下一篇

一种基于GMM的石质文物点云病害检测方法

张嘉敏1, 崔昊1, 吕红医2, 焦建辉3   

  1. 1. 郑州大学 地球科学与技术学院, 河南 郑州 450001;
    2. 龙门石窟研究院, 河南 洛阳 471023;
    3. 郑州大学 建筑学院, 河南 郑州 450001
  • 收稿日期:2026-02-02 发布日期:2026-08-01
  • 通信作者: 崔昊，副教授，研究方向为点云处理算法、深度学习、可解释机器学习。E-mail:cuihao@zzu.edu.cn E-mail:cuihao@zzu.edu.cn
  • 基金资助:
    国家自然科学基金(No.42241759)

A Disease Detection Method for Point Cloud of Stone Cultural Relics Based on GMM

ZHANG Jiamin1, CUI Hao1, LYU Hongyi2, JIAO Jianhui3   

  1. 1. School of Geo-Science and Technology, Zhengzhou University, Zhengzhou 450001, Henan, China;
    2. Longmen Grottoes Academy, Luoyang 471023, Henan, China;
    3. School of Architecture, Zhengzhou University, Zhengzhou 450001, Henan, China
  • Received:2026-02-02 Published:2026-08-01

PDF

6

摘要： 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。

关键词: 石质文物保护, 三维点云, 病害检测, 高斯混合模型, 局部凸包连接片

Abstract: 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.

Key words: stone cultural relic conservation, three-dimensional point cloud, damage detection, Gaussian mixture model, locally convex connected patch

中图分类号: