| 基于点云特征提取的薄壁文物碎片配对研究 |
| |
| 引用本文： | 刘恩盛,罗春泳,程小龙,黄玉花,戴维康,张宏.基于点云特征提取的薄壁文物碎片配对研究[J].井冈山大学学报（自然科学版）,2023,44(5):86-92. |
| |
| 作者姓名： | 刘恩盛 罗春泳 程小龙 黄玉花 戴维康 张宏 |
| |
| 作者单位： | 井冈山大学建筑工程学院, 江西, 吉安 343009;江西理工大学土木与测绘学院, 江西, 赣州 341000;上海博物馆, 上海, 200003;江西省地质局第九地质大队, 江西, 南昌 330000 |
| |
| 基金项目： | 国家自然科学基金项目(41662018)；江西省高校人文社会科学研究2022年度项目(KG22101)；江西省自然科学基金青年项目(20224BAB212025)；吉安市指导性科技计划项目(吉市科计字[2022]6号21) |
| |
| 摘 要： | 针对考古发掘现场堆积在一起的陶瓷碎片难以修复拼接，以及薄壁碎片断裂面特征难以提取的问题，提出了一种综合碎片点云多种特征提取与粗糙集模型进行约简分类的方法。该方法先采用最小半径移动球和三维中心滤波圆弧拟合的方法，分别获取碎片边缘和内部几何特征，再结合颜色纹理及材质等特征要素，通过决策表约简分类为文物碎片重建匹配提供条件。与现有依赖对象单一特征的配对方法不同，本方法在碎片存在磨损甚至缺失等情形下，能将混杂在一起的隶属于多器物薄壁文物碎片进行有效分类配对。
|
| 关 键 词： | 薄壁文物碎片 特征提取 三维中心滤波圆弧拟合 粗糙集模型 |
| 收稿时间： | 2023/4/4 0:00:00 |
| 修稿时间： | 2023/7/25 0:00:00 |
STUDY ON THIN WALL CULTURAL RELIC FRAGMENT PAIRING BASED ON POINT CLOUD FEATURE EXTRACTION |
| |
| LIU En-sheng,LUO Chun-yong,CHENG Xiao-long,HUANG Yu-hu,DAI Wei-kang,ZHANG Hong.STUDY ON THIN WALL CULTURAL RELIC FRAGMENT PAIRING BASED ON POINT CLOUD FEATURE EXTRACTION[J].Journal of Jinggangshan University(Natural Sciences Edition),2023,44(5):86-92. |
| |
| Authors: | LIU En-sheng LUO Chun-yong CHENG Xiao-long HUANG Yu-hu DAI Wei-kang ZHANG Hong |
| |
| Affiliation: | School of Building Engineering, Jinggangshan University, Ji''an, Jiangxi 343009, China;College of Civil Engineering and Surveying, Jiangxi University of Science and Technology, Ganzhou, Jiangxi 343000, China;Shanghai Museum, Shanghai 200003, China; The Ninth Geological Brigade of Jiangxi Provincial Geological Bureau, Nanchang, Jiangxi 330000, China |
| |
| Abstract: | In view of the problems that the ceramic fragments piled together in the archaeological excavation site are difficult to repair and splice, and the fracture surface characteristics of the thin-wall fragments are difficult to be extracted, in this paper, a method of reducing and classifying based on multiple features extraction of debris point cloud and rough set model is proposed. Firstly, the minimum radius moving sphere and the 3d central filtering arc fitting method are used to obtain the edge and internal geometric features of debris, respectively, combining with color, texture and material features, the decision table can be used to reduce the classification and provide the conditions for the reconstruction and matching of cultural relic fragments. Compared with the existing single feature matching method, this method can effectively classify and match the debris belonging to multi-objects under the conditions of debris wear or even missing. |
| |
| Keywords: | thin wall artifacts feature extraction 3D central filtering arc fitting rough set model |
|
| 点击此处可从《井冈山大学学报（自然科学版）》浏览原始摘要信息 |
|
点击此处可从《井冈山大学学报（自然科学版）》下载全文 |
|
