改进Chan-Vese模型的电力设备红外图像分割算法

张秋铭; 李云红; 罗雪敏; 屈海涛; 苏雪平; 任劼; 周小计

改进Chan-Vese模型的电力设备红外图像分割算法

张秋铭^{1, 2},
李云红^1, ,,
罗雪敏¹,
屈海涛³,
苏雪平¹,
任劼¹,
周小计¹

1.
西安工程大学电子信息学院, 陕西西安 710048
2.
国家电网四川省电力公司信息通信公司, 四川成都 610000
3.
枣庄学院, 山东枣庄 277160

基金项目:

国家自然科学基金 61902301

陕西省科技厅自然科学基础研究重点项目 2022JZ-35

陕西省教育厅自然科学基础研究计划 19JK0364

详细信息

作者简介:
张秋铭（1995-），女，助理工程师，研究方向是图像处理、计算机视觉

通讯作者:
李云红（1974-），女，教授，研究方向为红外热像技术、数字图像处理和信号与信息处理技术。E-mail：hitliyunhong@163.com

中图分类号: TP751.1
计量
- 文章访问数: 140
- HTML全文浏览量: 54
- PDF下载量: 42
- 被引次数: 0
出版历程
- 收稿日期: 2021-10-07
- 修回日期: 2021-11-10
- 刊出日期: 2023-02-20

Electric Equipment Infrared Image Segmentation Method Based on Improved Chan-Vese Model

1.
School of Electronics and Information, Xi'an Polytechnic University, Xi'an 710048, China
2.
Information Communication Company of Sichuan State Grid, Chengdu 610000, China
3.
Zaozhuang University, Zaozhaung 277160, China

摘要

摘要: 针对电力设备在线监测系统中红外图像分割效果差，速度慢等问题，提出一种改进的Chan-Vese模型的红外图像分割算法。首先，通过引入边缘能量项，一方面增强模型的局部控制能力，另一方面有效抑制了轮廓偏移。其次，利用径向基函数取代了传统的长度正则项，简化了计算。然后，通过引入内部能量项省去初始化过程，节省了算法的运行时间。经实验验证，Dice重合率（Dice similarity coefficient, DSC）平均值为0.9808，错误分割率（ratio of segmentation error, RSE）平均值为0.025，算法运行时间比其他模型总体平均值低66.8%。改进后的Chan-Vese模型分割算法的Dice重合率和错误分割率等均优于GAC-CV、CV-RSF、区域型水平集和Multiphase-CV模型分割算法。
- 红外图像分割 /
- Chan-Vese模型 /
- 边缘能量项 /
- 径向基函数
Abstract: To address the problems of poor infrared image segmentation and slow speed in the online monitoring system of power equipment, an improved infrared image segmentation algorithm based on the Chan-Vese model is proposed. First, by introducing the edge energy term, the local control ability of the model is enhanced and the contour shift is effectively suppressed. Second, a radial basis function is used to replace the traditional length regularization term, which simplifies the calculation. Subsequently, the initialization process is omitted by introducing internal energy items, which reduces the running time of the algorithm. After the experimental verification, the average DSC was 0.9808, the average value was 0.025, and the algorithm running time was 66.8% lower than the overall average of the other models. The improved Chan-Vese model segmentation algorithms DSC and RSE are better than the GAC-CV, CV-RSF, regional level set, and multiphase-CV model segmentation algorithms.
- infrared image segmentation /
- Chan-Vese model /
- edge energy term /
- radial basis function

HTML全文

图 1 Chan-Vese算法示意图

Figure 1. Schematic diagram of Chan-Vese algorithm

下载: 全尺寸图片幻灯片

图 2 轮廓偏移示意图

Figure 2. Schematic diagram of contour offset

下载: 全尺寸图片幻灯片

图 3 改进后Chan-Vese算法流程

Figure 3. Improved Chan-Vese algorithm flow chart

下载: 全尺寸图片幻灯片

图 4 管芯电阻分割对比

Figure 4. Die resistance segmentation comparison chart

下载: 全尺寸图片幻灯片

图 5 低压柜分割对比

Figure 5. Low-voltage cabinet segmentation comparison chart

下载: 全尺寸图片幻灯片

图 6 交直流接触器分割对比

Figure 6. AC and DC contactor segmentation comparison chart

下载: 全尺寸图片幻灯片

图 7 电容器组分割对比

Figure 7. Capacitor bank segmentation comparison chart

下载: 全尺寸图片幻灯片

图 8 电压互感器分割对比

Figure 8. Voltage transformer segmentation comparison chart

下载: 全尺寸图片幻灯片

图 9 中性点分割对比

Figure 9. Neutral point segmentation comparison chart

下载: 全尺寸图片幻灯片

表 1 DSC和RSE对比

Table 1. DSC and RSE comparison

	Fig.4		Fig.5		Fig.6		Fig.7		Fig.8		Fig.9
	DSC	RSE	DSC	RSE	DSC	RSE	DSC	RSE	DSC	RSE	DSC	RSE
Multiphase-CV	0.8754	0.0487	0.4383	0.5066	0.9733	0.0331	0.9123	0.1120	0.8532	0.1254	0.5031	0.4833
GAC-CV	0.8534	0.0425	0.6542	0.4524	0.9874	0.0324	0.9588	0.0421	0.4435	0.5673	0.9322	0.0695
CV-RSF	0.9789	0.0235	0.8821	0.3802	0.9325	0.0728	0.9614	0.0332	0.0233	0.8532	0.9475	0.0432
Region-based	0.9821	0.0232	0.3532	0.4244	0.9877	0.0242	0.9655	0.0310	0.9322	0.0614	0.9217	0.0782
Improved CV	0.9842	0.0217	0.9877	0.0232	0.9883	0.0237	0.9632	0.0315	0.9644	0.0223	0.9723	0.0334

下载: 导出CSV

表 2 分割时间对比

Table 2. Split time comparison s

	Fig.4	Fig.5	Fig.6	Fig.7	Fig.8	Fig.9
Multiphase-CV	20.7789	15.7865	3.2275	10.7602	8.5132	6.7832
GAC-CV	10.8322	15.5173	0.3688	7.8112	7.3345	3.2342
CV-RSF	9.2303	0.1344	0.2376	10.3325	67.5434	5.2874
Region-based	7.2304	30.5542	0.0834	12.4723	9.3567	5.8723
Improved CV	7.0322	0.0831	0.0886	5.6632	3.2127	2.7545

下载: 导出CSV

参考文献(21)

[1]	赵洪山, 刘秉聪, 王龄婕, 等. 基于压缩感知的电力设备红外图像盲超分辨率方法[J]. 电网技术, 2022, 46(3): 1177-1187. https://www.cnki.com.cn/Article/CJFDTOTAL-DWJS202203036.htm ZHAO H S, LIU B C, WANG L J, et al. Blind super resolution method for infrared image of power equipment based on compressed sensing[J]. Power System Technology, 2022, 46(3): 1177-1187. https://www.cnki.com.cn/Article/CJFDTOTAL-DWJS202203036.htm
[2]	曾军, 王东杰, 范伟, 等. 基于红外热成像的电气设备组件识别研究[J]. 红外技术, 2021, 43(7): 679-687. http://hwjs.nvir.cn/article/id/58024112-8052-43d6-8a2d-dd2460dfa5e1 ZENG J, WANG D J, FAN W, et al. Research on component identification for electrical equipment based on infrared thermography[J]. Infrared Technology, 2021, 43(7): 679-687. http://hwjs.nvir.cn/article/id/58024112-8052-43d6-8a2d-dd2460dfa5e1
[3]	李志坚, 杨风暴, 高玉斌, 等. 基于多尺度高斯滤波和形态学变换的红外与其他类型图像融合方法(英文)[J]. 红外与毫米波学报, 2020, 39(6): 810-817. https://www.cnki.com.cn/Article/CJFDTOTAL-HWYH202006021.htm LI Z J, YANG F B, GAO Y B, et al. Fusion method for infrared and other-type images based on the multi-scale Gaussian filtering and morphological transform[J]. Journal of Infrared and Millimeter Waves, 2020, 39(6): 810-817. https://www.cnki.com.cn/Article/CJFDTOTAL-HWYH202006021.htm
[4]	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.htm WANG X F, HU F K, HUANG S. 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
[5]	邵磊, 张一鸣, 李季, 等. 基于改进的两维Otsu管道红外图像高温区域分割研究(英文)[J]. 光谱学与光谱分析, 2019, 39(5): 1637-1642. https://www.cnki.com.cn/Article/CJFDTOTAL-GUAN201905055.htm SHAO L, ZHANG Y M, LI J, et al. Research on high temperature region segmentation of infrared pipeline image based on improve two-dimensional-Otsu[J]. Spectroscopy and Spectral Analysis, 2019, 39(5): 1637-1642. https://www.cnki.com.cn/Article/CJFDTOTAL-GUAN201905055.htm
[6]	聂方彦, 李建奇, 张平凤, 等. 图像非广延高斯熵分割的阈值选取方法[J]. 激光与红外, 2020, 50(12): 1504-1511. https://www.cnki.com.cn/Article/CJFDTOTAL-JGHW202012014.htm NIE F Y, LI J Q, ZHANG P F, et al. Threshold selection for image segmentation based on nonextensive Gaussian entropy[J]. Laser & Infrared, 2020, 50(12): 1504-1511. https://www.cnki.com.cn/Article/CJFDTOTAL-JGHW202012014.htm
[7]	朱莉, 张晶, 傅应锴, 等. 基于多模态特征图融合的红外热图像目标区域提取算法[J]. 红外与毫米波学报, 2019, 38(1): 125-132. https://www.cnki.com.cn/Article/CJFDTOTAL-HWYH201901020.htm ZHU L, ZHANG J, FU Y K, et al. Infrared thermal image ROI extraction algorithm based on fusion of multi-modal feature maps[J]. Journal of Infrared and Millimeter Waves, 2019, 38(1): 125-132. https://www.cnki.com.cn/Article/CJFDTOTAL-HWYH201901020.htm
[8]	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. https://www.cnki.com.cn/Article/CJFDTOTAL-HWJS202009007.htm LI T B, HHU J H, ZHOU Q K. Improved Moth-flame optimization algorithm based on Lévy flight to optimize infrared image segmentation[J]. Infrared Technology, 2020, 42(9): 846-854. https://www.cnki.com.cn/Article/CJFDTOTAL-HWJS202009007.htm
[9]	王硕禾, 巩方超, 古晓东, 等. 基于特征融合的变电设备类型及故障识别算法研究[J]. 铁道学报, 2021, 43(4): 95-100. https://www.cnki.com.cn/Article/CJFDTOTAL-TDXB202104014.htm WANG S H, GONG F C, GU X D et al. Research on types of substation equipment and fault identification algorithm based on feature fusion[J]. Journal of the China Railway Society, 2021, 43(4): 95-100. https://www.cnki.com.cn/Article/CJFDTOTAL-TDXB202104014.htm
[10]	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): 109-116. https://www.cnki.com.cn/Article/CJFDTOTAL-GXXB201908014.htm ZHAO Q S, WANG Y Y, WANG X P, et al. Substation infrared image segmentation based on novel threshold selection method[J]. Acta Optica Sinica, 2019, 39(8): 109-116. https://www.cnki.com.cn/Article/CJFDTOTAL-GXXB201908014.htm
[11]	于晓, 周子杰, Kamil Ríha. 基于最小平均距离免疫算法的模糊红外图像分割(英文)[J]. 光谱学与光谱分析, 2018, 38(11): 3645-3652. https://www.cnki.com.cn/Article/CJFDTOTAL-GUAN201811059.htm YU X, ZHOU Z J, Kamil Ríha. Blurred infrared image segmentation using new immune algorithm with minimum mean distance immune field[J]. Spectroscopy and Spectral Analysis, 2018, 38(11): 3645-3652. https://www.cnki.com.cn/Article/CJFDTOTAL-GUAN201811059.htm
[12]	Houssein Essam H, Helmy Bahaa El-din, Oliva Diego, et al. A novel black widow optimization algorithm for multilevel thresholding image segmentation[J]. Expert Systems With Applications, 2021, 167: 114159.
[13]	KotaridisIoannis, Lazaridou Maria. Remote sensing image segmentation advances: a meta-analysis[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2021, 173: 309-322.
[14]	周子杰, 张宝峰, 于晓. 基于神经免疫生长可免域网络的红外光谱图像分割算法(英文)[J]. 光谱学与光谱分析, 2021, 41(5): 1652-1660. https://www.cnki.com.cn/Article/CJFDTOTAL-GUAN202105059.htm ZHOU Z J, ZHANG B F, YU X. Infrared spectroscopic image segmentation based on neural immune network with growing immune field[J]. Spectroscopy and Spectral Analysis, 2021, 41(5): 1652-1660 https://www.cnki.com.cn/Article/CJFDTOTAL-GUAN202105059.htm
[15]	朱莉, 赵俊, 傅应锴, 等. 一种红外热图像目标区域分割的深度学习算法[J]. 西安电子科技大学学报, 2019, 46(4): 107-114, 121. https://www.cnki.com.cn/Article/CJFDTOTAL-XDKD201904015.htm ZHU L, ZHAO J, FU Y K, et al. Deep learning algorithm for the segmentation of the interested region of an infrared thermal image[J]. Journal of Xidian University, 2019, 46(4): 107-114, 121. https://www.cnki.com.cn/Article/CJFDTOTAL-XDKD201904015.htm
[16]	黄鸿, 吕容飞, 陶俊利, 等. 基于改进U-Net++的CT影像肺结节分割算法[J]. 光子学报, 2021, 50(2): 73-83. https://www.cnki.com.cn/Article/CJFDTOTAL-GZXB202102009.htm HUANG H, LV R F, TAO J L, et al. CT image lung nodule segmentation algorithm based on improved U-Net++[J]. Acta Photonica Sinica, 2021, 50(2): 73-83. https://www.cnki.com.cn/Article/CJFDTOTAL-GZXB202102009.htm
[17]	王珏, 张秀英, 蔡玉芳, 等. 联合小波变换和RSF模型的CT图像分割方法[J]. 光学学报, 2020, 40(21): 57-65. https://www.cnki.com.cn/Article/CJFDTOTAL-GXXB202021007.htm WANG J, ZHANG X Y, CAI Y F, et al. CT image segmentation method combining wavelet transform and RSF model[J]. Acta Optica Sinica, 2020, 40(21): 57-65. https://www.cnki.com.cn/Article/CJFDTOTAL-GXXB202021007.htm
[18]	Aldísio G Medeiros, Matheus T Guimarães, Solon A Peixoto, et al. A new fast morphological geodesic active contour method for lung CT image segmentation[J]. Measurement, 2019, 148: 1-13.
[19]	Chan T F, Vese L A. Active contours without edges[J]. IEEE Transactions on Image Processing, 2001, 10(2): 266-277.
[20]	何婷, 郭艳光, 辛春花. ACM选择系统结合改进Chan-Vese模型的自适应图像分割方法[J]. 重庆理工大学学报(自然科学), 2020, 34(2): 165-173. https://www.cnki.com.cn/Article/CJFDTOTAL-CGGL202002024.htm HE T, GUO Y G XIN C H. Adaptive image segmentation method using ACM selection system and improved Chan-Vese model[J]. Journal of Chongqing University of Technology(Natural Science), 2020, 34(2): 165-173. https://www.cnki.com.cn/Article/CJFDTOTAL-CGGL202002024.htm
[21]	赵晓理, 周浦城, 薛模根. 一种基于改进Chan-Vese模型的红外图像分割方法[J]. 红外技术, 2016, 38(9): 774-778. http://hwjs.nvir.cn/article/id/hwjs201609011 ZHAO X L, ZHOU P C, XUE M G. A kind of infrared image segment method using improved Chan-Vese model[J]. Infrared Technology, 2016, 38(9): 774-778. http://hwjs.nvir.cn/article/id/hwjs201609011