留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于模糊推理的电气设备红外图像分割

曾水玲 唐敏之

曾水玲, 唐敏之. 基于模糊推理的电气设备红外图像分割[J]. 红外技术, 2023, 45(5): 446-454.
引用本文: 曾水玲, 唐敏之. 基于模糊推理的电气设备红外图像分割[J]. 红外技术, 2023, 45(5): 446-454.
ZENG Shuiling, TANG Minzhi. Infrared Image Segmentation for Electrical Equipment based on Fuzzy Inference[J]. Infrared Technology , 2023, 45(5): 446-454.
Citation: ZENG Shuiling, TANG Minzhi. Infrared Image Segmentation for Electrical Equipment based on Fuzzy Inference[J]. Infrared Technology , 2023, 45(5): 446-454.

基于模糊推理的电气设备红外图像分割

基金项目: 

国家自然科学基金项目 61363033

国家自然科学基金项目 61966014

湖南省研究生科研创新项目 CX20221107

江苏省社会安全图像与视频理解重点实验室开放课题 202212

吉首大学校级科研项目 Jdy22026

详细信息
    作者简介:

    曾水玲(1975-),女,博士,教授,研究方向为人工智能与模式识别。E-mail: zengflsl@163.com

  • 中图分类号: TP391

Infrared Image Segmentation for Electrical Equipment based on Fuzzy Inference

  • 摘要: 使用模糊理论处理电气设备红外图像分割的不确定性,提出了一种基于模糊推理的电气设备红外图像分割算法。首先分别利用电气设备红外图像故障区域的像素灰度、像素点与图像质心的马氏距离以及图像膨胀操作定义了强度特征、全局故障可能性特征和局部灰度特征;然后根据特征的模糊语言值制定了27条模糊规则,设计了一种模糊推理红外图像分割算法;最后,从主观和客观评价指标上将算法与传统Otsu算法和FCM算法进行了对比。实验表明,该算法的分割精度和误分割率比其他两种算法都有一定的改善,同时该算法能够滤除图像中具有高亮度的干扰区域,对具有小亮度差和小面积故障区域的红外图像有较好的分割效果。
  • 图  1  模糊推理系统结构图

    Figure  1.  Structure diagram of fuzzy inference system

    图  2  强度特征隶属度函数

    Figure  2.  Intensity feature membership function

    图  3  马氏距离等高线

    Figure  3.  Mahalanobis distance contour

    图  4  欧氏距离等高线

    Figure  4.  Euclidean distance contour

    图  5  全局故障可能性特征隶属度函数

    Figure  5.  Global failure probability feature membership function

    图  6  局部灰度分布特征隶属度函数

    Figure  6.  Local gray distribution features membership function

    图  7  基于模糊推理的电气设备红外图像分割算法流程图

    Figure  7.  Flow chart of infrared image segmentation algorithm for electrical equipment based on fuzzy reasoning

    图  8  110 kV变电站红外图像

    Figure  8.  Infrared image of 110kV substation

    图  9  110 kV变电站红外灰度图像

    Figure  9.  Infrared gray level image of 110kV substation

    图  10  可能故障区域

    Figure  10.  Possible failure area

    图  11  强度特征图

    Figure  11.  Intensity characteristic map

    图  12  全局故障可能性特征图

    Figure  12.  Global fault possibility feature diagram

    图  13  局部灰度分布特征图

    Figure  13.  Local gray distribution feature

    图  14  模糊推理输出图像图

    Figure  14.  Fuzzy inference output image

    图  15  分割图像

    Figure  15.  Segmented image

    图  16  电气设备红外图像

    Figure  16.  Infrared image of electrical equipment

    图  17  Otsu算法分割效果

    Figure  17.  Segmentation effect of Otsu algorithm

    图  18  FCM算法分割效果

    Figure  18.  Segmentation effect of FCM algorithm

    图  19  本文算法的分割效果

    Figure  19.  The segmentation effect of the algorithm in this paper

    图  20  Ground Truth标准分割图像

    Figure  20.  Ground Truth standard segmentation image

    表  1  基于多特征改进的模糊规则

    Table  1.   Fuzzy rules based on multi-feature improvement

    Intensity characteristic Global failure possibility characteristics Local gray distributioncharacteristics Failure membership characteristics
    1 small small small small
    2 small small medium small
    3 small small big medium
    4 small medium small small
    5 small medium medium medium
    6 small medium big medium
    7 small big small small
    8 small big medium small
    9 small big big big
    10 medium small small medium
    11 medium small medium medium
    12 medium small big medium
    13 medium medium small medium
    14 medium medium medium medium
    15 medium medium big big
    16 medium big small medium
    17 medium big medium medium
    18 medium big big big
    19 big small small medium
    20 big small medium medium
    21 big small big medium
    22 big medium small medium
    23 big medium medium medium
    24 big medium big big
    25 big big small medium
    26 big big medium medium
    27 big big big big
    下载: 导出CSV

    表  2  算法评价指标对比

    Table  2.   Algorithm evaluation index comparison

    Figure 1 Figure 2 Figure 3 Figure 4
    IOU ME IOU ME IOU ME IOU ME
    Otsu algorithm 0.59% 49.30% 0.30% 21.67% 83.55% 1.70% 25.64% 11.73%
    FCM algorithm 1.52% 19.04% 0.56% 11.44% 91.60% 0.77% 91.27% 0.39%
    Algorithm two in this paper 90.02% 0.03% 95.00% 0.01% 92.50% 0.67% 97.60% 0.10%
    下载: 导出CSV
  • [1] 苏海锋, 赵岩, 武泽君, 等. 基于改进RetinaNet的电力设备红外目标精细化检测模型[J]. 红外技术, 2021, 43(11): 1104-1111. http://hwjs.nvir.cn/article/id/3233a6a1-cbf0-4110-baa5-2a56e551f092

    SU Haifeng, ZHAO Yan, WU Zejun, et al. Refined infrared object detection model for power equipment based on improved RetinaNet[J]. Infrared Technology, 2021, 43(11): 1104-1111. http://hwjs.nvir.cn/article/id/3233a6a1-cbf0-4110-baa5-2a56e551f092
    [2] 陈飞. 改进的交互式Otsu红外图像分割算法[J]. 计算机测量与控制, 2020, 28(9): 248- 251. https://www.cnki.com.cn/Article/CJFDTOTAL-JZCK202009049.htm

    CHEN Fei. An improved interactive Otsu infrared image segmentation algorithm[J]. Computer Measurement & Control, 2020, 28(9): 248-251. https://www.cnki.com.cn/Article/CJFDTOTAL-JZCK202009049.htm
    [3] YU X, ZHOU Z, GAO Q, et al. Infrared image segmentation using growing immune field and clone threshold[J]. Infrared Physics & Technology, 2018, 88: 184-193. http://www.sciencedirect.com/science/article/pii/S1350449517306308
    [4] 冯振新, 周东国, 江翼, 等. 基于改进MSER算法的电力设备红外故障区域提取方法[J]. 电力系统保护与控制, 2019, 47(5): 123-128. https://www.cnki.com.cn/Article/CJFDTOTAL-JDQW201905015.htm

    FENG Zhenxin, ZHOU Dongguo, JIANG Yi, et al. Fault region extraction using improved MSER algorithm with application to the electrical system[J]. Power System Protection and Control, 2019, 47(5): 123-128. https://www.cnki.com.cn/Article/CJFDTOTAL-JDQW201905015.htm
    [5] HU F, CHEN H, WANG X. An intuitionistic Kernel-based fuzzy C-means clustering algorithm with local information for power equipment image segmentation[J]. IEEE Access, 2020, 8: 4500-4514. doi:  10.1109/ACCESS.2019.2963444
    [6] QI C, LI Q, LIU Y, et al. Infrared image segmentation based on multi-information fused fuzzy clustering method for electrical equipment[J]. International Journal of Advanced Robotic Systems, 2020, 17(2): 1729881420909600. http://www.xueshufan.com/publication/3009600905
    [7] WEI D, WANG Z, SI L, et al. An image segmentation method based on a modified local-information weighted intuitionistic Fuzzy C-means clustering and gold-panning algorithm[J]. Engineering Applications of Artificial Intelligence, 2021, 101: 104209. doi:  10.1016/j.engappai.2021.104209
    [8] BAI X, LIU M, WANG T, et al. Feature based fuzzy inference system for segmentation of low-contrast infrared ship images[J]. Applied Soft Computing, 2016, 46: 128-142. doi:  10.1016/j.asoc.2016.05.004
    [9] 王力斌, 刘树伟. 基于模糊推理的油门防误踩系统控制研究[J]. 控制工程, 2020, 27(8): 1462-1467. https://www.cnki.com.cn/Article/CJFDTOTAL-JZDF202008026.htm

    WANG Libin, LIU Shuwei. Research on the control system of preventing false stepping the accelerating pedal based on the fuzzy theory[J]. Control Engineering of China, 2020, 27(8): 1462-1467. https://www.cnki.com.cn/Article/CJFDTOTAL-JZDF202008026.htm
    [10] 张洪群, 顾吟雪, 郭擎. 灰色关联分析与模糊推理边缘检测图像融合法[J]. 遥感信息, 2020, 35(1): 15-27. https://www.cnki.com.cn/Article/CJFDTOTAL-YGXX202001003.htm

    ZHANG Hongqun, GU Yinxue, GUO Qing. Image fusion based edge detection of grey relational analysis and fuzzy inference[J]. Remote Sensing Information, 2020, 35(1): 15-27. https://www.cnki.com.cn/Article/CJFDTOTAL-YGXX202001003.htm
    [11] Vijayakumar S, Santhi V. Speckle noise reduction in SAR images using fuzzy inference system[J]. International Journal of Fuzzy System Applications (IJFSA), 2019, 8(4): 60-83.
    [12] Arya K V. A new fuzzy rule based pixel organization scheme for optimal edge detection and impulse noise removal[J]. Multimedia Tools and Applications, 2020, 79(45): 33811-33837. doi:  10.1007/s11042-020-08707-x?utm_source=xmol&utm_content=meta
    [13] BAI X, CHEN Z, ZHANG Y, et al. Infrared ship target segmentation based on spatial information improved FCM[J]. IEEE Transactions On Cybernetics, 2015, 46(12): 3259-3271. http://ieeexplore.ieee.org/document/7026038/
    [14] LONG J, Shelhamer E, Darrell T. Fully convolutional networks for semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2015: 3431-3440.
  • 加载中
图(20) / 表(2)
计量
  • 文章访问数:  137
  • HTML全文浏览量:  22
  • PDF下载量:  23
  • 被引次数: 0
出版历程
  • 收稿日期:  2022-03-30
  • 修回日期:  2022-05-11
  • 刊出日期:  2023-05-20

目录

    /

    返回文章
    返回