基于机器视觉古陶瓷无损分类识别

发布时间：2018-01-30 12:27

本文关键词： 古陶瓷科技鉴定机器视觉结构信息釉色信息纹饰特征　出处：《硅酸盐学报》2017年12期 　论文类型：期刊论文

【摘要】：为客观、有效地对古陶瓷进行无损分类,提出了一种基于机器视觉古陶瓷无损分类识别方法。通过遍历古陶瓷器型边缘轮廓,获取古陶瓷器型结构细节特征,并在HSI空间下提取古陶瓷釉色多通道颜色直方图特征。同时,提取反映古陶瓷纹理多样性的LBP纹饰特征。基于上述特征,采用机器学习方法实现古陶瓷器型结构、釉色及其纹饰图案的无损分类识别。结果表明:通过机器视觉可以有效地对古陶瓷进行分类识别;在以16为曲率步长、9为LBP算子分块数时,分别提取古陶瓷结构,纹饰特征有较好的识别精度,其中,基于结构与釉色融合特征相比单一特征具有更好的识别效果;当古陶瓷发生结构或纹饰上的小部分缺损时,该方法可以保持一定的鲁棒性,当信息丢失或缺损为5%时,平均识别率依旧可达85%以上,可期望实现古陶瓷科技鉴定中的良好应用。
[Abstract]:In order to classify ancient ceramics objectively and effectively, a new method based on machine vision was proposed. By traversing the edge contours of ancient ceramics, the structural details of ancient ceramics were obtained. The multi-channel color histogram features of ancient ceramic glaze were extracted in HSI space. At the same time, LBP decorative features reflecting the texture diversity of ancient ceramics were extracted, based on the above features. The machine learning method is used to realize the nondestructive classification and recognition of the structure, glaze and patterns of ancient ceramics. The results show that the classification and recognition of ancient ceramics can be effectively carried out by machine vision. With 16 as the curvature step size and 9 as the LBP operator block number, the ancient ceramic structures are extracted, and the decorative features have good recognition accuracy. Compared with single feature, the recognition effect based on structure and glaze fusion is better than that of single feature. The method can maintain a certain robustness when a small part of the defect on the structure or decoration occurs. When the information is lost or the defect is 5, the average recognition rate can still reach more than 85%. It can be expected to realize the good application in the scientific and technological appraisal of ancient ceramics.
【作者单位】：上海大学通信与信息工程学院;新型显示技术及应用集成教育部重点实验室;上海大学材料科学与工程学院;
【基金】：国家自然科学基金(11176016;60872117) 高等学校博士学科点专项科研基金(20123108110014)资助
【分类号】：TP391.41;TQ174.66
【正文快照】： 2.新型显示技术及应用集成教育部重点实验室,上海200072;3.上海大学材料科学与工程学院,上海200444)WENG Zhengkui1,GUAN Yepeng1,2,LUO Hongjie3(1.School of Communication and Information Engineering,Shanghai University,Shanghai 200444,China;2.Key Laboratory of Adv

【相似文献】