Infrared Image Segmentation Method Based on Fuzzy Clustering with Similarity Thresholding
-
摘要: 针对输电线路电力设备红外图像热故障区域检测,提出采用一种基于相似度阈值的模糊聚类热故障区域提取方法。在该方法中,改进了传统模糊均值聚类算法的迭代求解方式,采用一种阈值化模糊聚类;其次,通过对目标区域局部邻域像素的相似度聚类分析,并结合其隶属度的计算,确保局部邻域像素在聚类上的相似性。同时,引入了最大相似度阈值准则简化均值的设置以及自高向低的迭代方式,从而提升区域提取效率。最后通过真实输电线路电气设备红外故障图像测试,验证了文中所提方法的有效性和适用性。Abstract: This paper presents a fuzzy clustering method based on similarity thresholding to detect an overheating fault region from an infrared image of a transmission line. In this method, the original iteration mechanism of fuzzy clustering was improved and a thresholding fuzzy clustering model was built. Thus, a fuzzy member was utilized to measure the neighboring pixels t by conducting cluster analysis on the object region with local neighboring pixels. This ensured similarity during the clustering of the local neighboring pixels into the cluster center. In addition, the maximum similarity thresholding rule was applied to determine the final thresholding using the strategy of thresholding from top to bottom, thus improving the efficiency of the method in obtaining the final region of interest in the infrared image using fuzzy clustering. Finally, experimental results on infrared images of transmission lines show that the good performance of the proposed method and that the proposed method is suitable for fault detection in transmission lines.
-
Key words:
- similarity thresholding /
- fuzzy clustering /
- infrared image /
- neighboring pixels
-
表 1 Otsu和MST方法阈值
Table 1. Thresholding of MST and Otsu
Method Image 1 Image 2 Image 3 Image 4 Image 5 Otsu 21 59 31 57 53 MST 15 45 22 41 84 表 2 最终迭代后的均值
Table 2. Mean value of final results
Image 1 Image 2 Image 3 Image 4 Image 5 FCM v0 1.0603 21.05 1.01730 5.7834 24.3979 v1 41.3490 96.8447 53.8848 31.1222 84.1881 Ours v0 12.1560 65.2512 30.4860 14.9498 48.7644 v1 130.9086 133.1862 147.4263 129.3911 209.7107 表 3 相似度评价结果
Table 3. Evaluation of similarity
Image 1 Image 2 Image 3 Image 4 Image 5 MST 0.5120 0.3872 0.3433 0.4815 0.3552 Ours 0.3430 0.2419 0.2445 0.3515 0.1962 表 4 不同方法运行时长对比
Table 4. Comparison of time speeding methods in evaluation
s Image 1 Image 2 Image 3 Image 4 Image 5 Otsu 0.1402 0.0022 0.1076 0.0042 0.2027 MST 3.7802 0.2007 3.4117 0.9979 2.32 FCM 5.6357 0.9010 13.6103 7.7979 15.1982 Meanshift 5.5414 0.2870 7.5225 1.1907 12.3957 Ours 4.3923 0.2682 5.8280 0.6663 3.1632 -
[1] 王淼, 杜伟, 孙鸿博, 等. 基于红外图像识别的输电线路故障诊断方法[J]. 红外技术, 2017, 39(4): 383-386. http://hwjs.nvir.cn/article/id/hwjs201704015WANG Miao, DU Wei, SUN Hongbo, et al. Transmission line fault diagnosis method based on infrared image recognition[J]. Infrared Technology, 2017, 39(4): 383-386. http://hwjs.nvir.cn/article/id/hwjs201704015 [2] 张福, 张建刚, 李庭坚, 等. 基于红外热点的电力缺陷检测方法[J]. 电网与清洁能源, 2018, 34(3): 46-50. doi: 10.3969/j.issn.1674-3814.2018.03.009ZHANG Fu, ZHANG Jiangang, LI Tingjian, et al. An electrical power fault detection method based on infrared image heating[J]. Power System and Clean Energy, 2018, 34(3): 46-50. doi: 10.3969/j.issn.1674-3814.2018.03.009 [3] 刘云鹏, 裴少通, 武建华. 基于深度学习的输变电设备异常发热点红外图片目标检测方法[J]. 南方电网技术, 2019, 13(2): 27-33. https://www.cnki.com.cn/Article/CJFDTOTAL-NFDW201902006.htmLIU Yunpeng, PEI Shaotong, WU Jianhua. Deep learning based target detection method for abnormal hot spots infrared images of transmission and transformation equipment[J]. Southern Power System Technology, 2019, 13(2): 27-33. https://www.cnki.com.cn/Article/CJFDTOTAL-NFDW201902006.htm [4] 白梓璇, 高强, 于晓, 等. 基于补体免疫聚类的电力设备红外图像目标提取算法[J]. 激光杂志, 2020, 41(3): 107-112. https://www.cnki.com.cn/Article/CJFDTOTAL-JGZZ202003021.htmBAI Zixuan, GAO Qiang, YU Xiao, et al. Infrared image target extraction method for power equipment based on complement immune clustering[J]. Laser Journal, 2020, 41(3): 107-112. https://www.cnki.com.cn/Article/CJFDTOTAL-JGZZ202003021.htm [5] 井金剑, 翟春艳, 李书臣, 等. 基于红外图像识别的电气设备故障诊断[J]. 电子设计工程, 2014, 22(12): 171-173. doi: 10.3969/j.issn.1674-6236.2014.12.059JING Jinjian, ZHAI Chunyan, LI Shuchen, et al. Fault diagnosis of electrical equipment based on infrared image recognition[J]. Electronic Design Engineering, 2014, 22(12): 171-173. doi: 10.3969/j.issn.1674-6236.2014.12.059 [6] 赵庆生, 王雨滢, 王旭平, 等. 基于新型阈值选择方法的变电站红外图像分割[J]. 光学学报, 2019, 39(8): 101-108. https://www.cnki.com.cn/Article/CJFDTOTAL-GXXB201908014.htmZHAO Qingsheng, WANG Yuying, WANG Xuping, et al. Substation infrared image segmentation based on novel threshold selection method[J]. Acta Optica Sinica, 2019, 39(8): 101-108. https://www.cnki.com.cn/Article/CJFDTOTAL-GXXB201908014.htm [7] 余成波, 曾亮, 张林. 基于OTSU和区域生长的电气设备多点故障分割[J]. 红外技术, 2018, 40(10): 1008-1012. http://hwjs.nvir.cn/article/id/hwjs201810013YU Chengbo, ZENG Liang, ZHANG Lin. Multi point fault segmentation of electrical equipment based on OTSU and region growth[J]. Infrared Technology, 2018, 40(10): 1008-1012. http://hwjs.nvir.cn/article/id/hwjs201810013 [8] 王启银, 薛建东, 任新辉. 一种自适应的变电站设备红外图像分割方法[J]. 红外技术, 2016, 38(9): 770-773. http://hwjs.nvir.cn/article/id/hwjs201609010WANG Qiyin, XUE Jiandong, REN Xinhui. An adaptive segmentation method of substation equipment infrared image[J]. Infrared Technology, 2016, 38(9): 770-773. http://hwjs.nvir.cn/article/id/hwjs201609010 [9] 李唐兵, 胡锦泓. 周求宽. 基于Lévy飞行的改进飞蛾扑火算法优化红外图像分割[J]. 红外技术, 2020, 42(9): 846-854. http://hwjs.nvir.cn/article/id/hwjs202009006LI Tangbing, HU Jinhong, ZHOU Qiukuan. Improved moth-flame optimization algorithm based on Lévy flight to optimize infrared image segmentation[J]. Infrared Technology, 2020, 42(9): 846-854. http://hwjs.nvir.cn/article/id/hwjs202009006 [10] 王小芳, 毛华敏. 一种复杂背景下的电力设备红外图像分割方法[J]. 红外技术, 2019, 41(12): 1111-1116. http://hwjs.nvir.cn/article/id/hwjs201912004WANG Xiaofang, MAO Huamin. Infrared image segmentation method for power equipment in complex background[J]. Infrared Technology, 2019, 41(12): 1111-1116. http://hwjs.nvir.cn/article/id/hwjs201912004 [11] 王晓飞, 胡凡奎, 黄硕. 基于分布信息直觉模糊c均值聚类的红外图像分割算法[J]. 通信学报, 2020, 41(5): 120-129. https://www.cnki.com.cn/Article/CJFDTOTAL-TXXB202005013.htmWANG Xiaofei, HU Fankui, HUANG Shuo. Infrared image segmentation algorithm based on distribution information intuitionistic fuzzy c-means clustering[J]. Journal on Communications, 2020, 41(5): 120-129. https://www.cnki.com.cn/Article/CJFDTOTAL-TXXB202005013.htm [12] 冯振新, 许晓路, 周东国, 等. 基于局部区域聚类的电力设备故障区域提取方法[J]. 电测与仪表, 2020, 57(8): 45-50. https://www.cnki.com.cn/Article/CJFDTOTAL-DCYQ202008008.htmFENG Zhenxin, XU Xiaolu, ZHOU Dongguo. Extraction method of power device fault region based on local clustering algorithm[J]. Electrical Measurement & Instrumentation, 2020, 57(8): 45-50. https://www.cnki.com.cn/Article/CJFDTOTAL-DCYQ202008008.htm [13] Mehmet Sezgin, Bülent Sankur. Survey over image thresholding techniques and quantitative performance evaluation[J]. Journal of Electronic Imaging, 2004, 13(1): 146-161. [14] ZOU Yaobin, DONG Fangmin, LEI Bangjun. Maximum similarity thresholding[J]. Digital Image Processing, 2014, 28: 120-135 [15] Bezdek J C. Pattern Recognition with Fuzzy Objective Function Algorithm[M]. New York: Springer, 1981. [16] Ahmed M N, Yamany S M, Mohamed N, et al. A modified fuzzy c-means algorithm for bias field estimation and segmentation of MRI data[J]. IEEE Transactions on Medical Imaging, 2002, 21: 193-199. doi: 10.1109/42.996338