|
|
|
| An Improved Monocular Ranging Method for Infrared Image of Power Equipment Based on the Pixel Width Recognition of Objects |
| Yang Fan1, Wang Mengjun1, Tan Tian1, Lu Xu2, Hu Ran2 |
1. The State Key Laboratory of Power Transmission Equipment & System Security and New Technology Chongqing University Chongqing 400044 China;
2. Electric Power Research Institute Shenzhen Power Supply Co. Ltd Shenzhen 518000 China |
|
|
|
Abstract The shooting distance is one of the main factors affecting the infrared imaging effect. Its accurate measurement is the main method to improve the fault accuracy of the infrared imaging detection equipment. This paper studies an improved algorithm of the infrared imaging monocular ranging of power equipment based on target pixel width recognition, aiming at the requirement of automatic ranging of the infrared thermal imager in the process of power equipment patrol inspection. Automatic distance recognition is realized by the pixel width of power equipment in the infrared image. The problem of complex distance recognition caused by infrared imaging of power equipment due to shooting angle transformation and incomplete equipment shooting is solved. The automatic distance recognition of 12 kinds of common power equipment is realized.
The device type in the image must be identified first to obtain the device's distance in the image. This paper automatically identifies the device type in the infrared image based on the single shot MultiBox detector (SSD) algorithm, and obtains the coordinates of the device type and identification frame. The average accuracy can reach 98.24 %. Since most of the power equipment is columnar in shape and large in size, to increase the proportion of equipment in the picture as much as possible when taking infrared images, the angle will tilt when taking pictures. As a result, the acquired images will have the following problems: (1) The whole equipment cannot be seen in the picture; (2) The pixel width of the target detection frame cannot represent the pixel width corresponding to the actual width of the device due to the oblique shooting angle. Therefore, this paper analyzes the characteristics of power infrared patrol inspection. It is found that even though the equipment angle is inclined, the overall appearance of the power equipment is cylindrical, and the maximum width of the equipment will not change. Even if the equipment is not photographed entirely, its width can still be reflected. Accordingly, the paper proposes an improved monocular ranging algorithm based on the target pixel width. The minimum adjacent rectangle of the equipment is recognized through image processing, the pixel width of the target is calculated, and the final output device type and distance. The results show that the average error of the 12 types of equipment is 0.257 m, the maximum recognition error is 0.398 m, and the average error rate is 1.31 %.
The experimental results show that the improved algorithm can meet the requirements of monocular distance measurement of power equipment infrared images. Based on the distance measurement method in this paper, the automatic distance recognition of the infrared intelligent diagnostic device of power equipment can be realized, and the correction of the imaging temperature can be realized through the relationship between the distance and the infrared imaging temperature. Therefore, the infrared detection accuracy can be improved, and the accurate infrared temperature measurement of power equipment can be realized.
|
|
Received: 26 January 2022
|
|
|
|
|
|
[1] 纽春萍, 矫璐璐, 王小华, 等. 基于多场耦合的环保型GIS热特性分析[J]. 电工技术学报, 2020, 35(17): 3765-3772.
Niu Chunping, Jiao Lulu, Wang Xiaohua, et al.Thermal characteristics analysis of environmentally friendly GIS based on multi-field coupling[J]. Transa- ctions of China Electrotechnical Society, 2020, 35(17): 3765-3772.
[2] Yu Xiao, Ye Xi, Gao Qiang.Infrared handprint image restoration algorithm based on apoptotic mecha- nism[J]. IEEE Access, 2020, 8: 47334-47343.
[3] 郑含博, 李金恒, 刘洋, 等. 基于改进YOLOv3的电力设备红外目标检测模型[J]. 电工技术学报, 2021, 36(7): 1389-1398.
Zheng Hanbo, Li Jinheng, Liu Yang, et al.Infrared object detection model for power equipment based on improved YOLOv3[J]. Transactions of China Elec- trotechnical Society, 2021, 36(7): 1389-1398.
[4] 徐奇伟, 黄宏, 张雪锋, 等. 基于改进区域全卷积网络的高压引线接头红外图像特征分析的在线故障诊断方法[J]. 电工技术学报, 2021, 36(7): 1380-1388.
Xu Qiwei, Huang Hong, Zhang Xuefeng, et al.Online fault diagnosis method for infrared image feature analysis of high-voltage lead connectors based on improved R-FCN[J]. Transactions of China Elec- trotechnical Society, 2021, 36(7): 1380-1388.
[5] 金立军, 张达, 段绍辉, 等. 基于红外与紫外图像信息融合的绝缘子污秽状态识别[J]. 电工技术学报, 2014, 29(8): 309-318.
Jin Lijun, Zhang Da, Duan Shaohui, et al.Recognition of contamination grades of insulators based on IR and UV image information fusion[J]. Transactions of China Electrotechnical Society, 2014, 29(8): 309-318.
[6] 张培铭, 江和, 李光辉, 等. 中压开关柜接头温度在线监测技术的研究[J]. 电工技术学报, 1995, 10(2): 49-52.
Zhang Peiming, Jiang He, Li Guanghui, et al.The study of on-line temperature detection technique for the contacts in MV switchgear cubicle[J]. Transa- ctions of China Electrotechnical Society, 1995, 10(2): 49-52.
[7] 胡红光. 电力设备红外诊断技术与应用[M]. 北京: 中国电力出版社, 2012.
[8] Faye E.Distance makes the difference in thermo- graphy for ecological studies[J]. Journal of Thermal Biology, 2016, 56: 1-9.
[9] 国家能源局. 带电设备红外诊断应用规范: DL/T 664—2016[S]. 北京: 中国电力出版社, 2016.
[10] 徐彪, 尹项根, 张哲, 等. 电网故障诊断的分阶段解析模型[J]. 电工技术学报, 2018, 33(17): 4113-4122.
Xu Biao, Yin Xianggen, Zhang Zhe, et al.A staged analytical model for power system fault diagnosis[J]. Transactions of China Electrotechnical Society, 2018, 33(17): 4113-4122.
[11] 李典阳, 张育杰, 冯健, 等. 变压器故障样本多维诊断及结果可信度分析[J]. 电工技术学报, 2022, 37(3): 667-675.
Li Dianyang, Zhang Yujie, Feng Jian, et al.Multi- dimensional diagnosis of transformer fault sample and credibility analysis[J]. Transactions of China Elec- trotechnical Society, 2022, 37(3): 667-675.
[12] 尹游, 周凯, 李诗雨, 等. 基于极化去极化电流法的水树老化XLPE电缆界面极化特性分析[J]. 电工技术学报, 2020, 35(12): 2643-2651.
Yin You, Zhou Kai, Li Shiyu, et al.Interface polarization characteristics of water tree aged XLPE cables based on polarization and depolarization current method[J]. Transactions of China Electro- technical Society, 2020, 35(12): 2643-2651.
[13] 陈涛, 刘志刚, 胡轲珽, 等. 一种双重化脉冲整流器多管开路故障快速诊断方法[J]. 电工技术学报, 2020, 35(10): 2226-2238.
Chen Tao, Liu Zhigang, Hu Keting, et al.Quick diagnosis method for double-PWM rectifier multi- tube open circuit fault[J]. Transactions of China Elec- trotechnical Society, 2020, 35(10): 2226-2238.
[14] Witus Gary, Hunt Shawm. Monocular visual ran- ging[J]. Proceedings of SPIE-The International Society for Optical Engineering, 2008, 6962: 696204(7).
[15] Zhao Chaoqiang, Sun Qiyu, Zhang Chongzhen, et al.Monocular depth estimation based on deep learning: an overview[J]. Science China Technological Sciences, 2020, 63(9): 1612-1627.
[16] Shen Chao, Zhao Xiangmo, Liu Zhanwen, et al.Joint vehicle detection and distance prediction via mono- cular depth estimation[J]. IET Intelligent Transport Systems, 2020, 14(7): 753-763.
[17] Liu Qiang, Pan Ming, Li Yongwei.Design of vehicle monocular ranging system based on FPGA[J]. Chinese Journal of Liquid Crystals and Displays, 2014, 29(3): 422-428.
[18] Chen Hao, Lin Meian, Xue Lixia, et al.Research on location method based on monocular vision[J]. Journal of Physics: Conference Series, 2021, 1961(1): 012063.
[19] 刘斌, 李港庆, 安澄全, 等. 基于多尺度特征融合的红外单目测距算法[J]. 计算机应用, 2022, 42(3): 804-809.
Liu Bin, Li Gangqing, An Chengquan, et al.Infrared monocular ranging algorithm based on multiscale feature fusion Chinese full text[J]. Journal of Com- puter Applications, 2022, 42(3): 804-809
[20] 韩延祥, 张志胜, 戴敏. 用于目标测距的单目视觉测量方法[J]. 光学精密工程, 2011, 19(5): 1110-1117.
Han Yanxiang, Zhang Zhisheng, Dai Min.Monocular vision system for distance measurement based on feature points[J]. Optics and Precision Engineering, 2011, 19(5): 1110-1117.
[21] 胡旻昊, 王海滨, 王岐, 等. 基于双目视觉图像的电力设备状态识别算法[J]. 电子世界, 2017(21): 16-18.
Hu Minhao, Wang Haibin, Wang Qi, et al.Power equipment state recognition algorithm based on binocular vision image[J]. Electronics World, 2017(21): 16-18.
[22] 董诗绘, 牛彩雯, 戴琨. 基于深度强化学习的变电站巡检机器人自动化控制方法研究[J]. 高压电器, 2021, 57(2): 172-177.
Dong Shihui, Niu Caiwen, Dai Kun.Study on automatic control method of substation inspection robot based on deep reinforcement learning[J]. High Voltage Apparatus, 2021, 57(2): 172-177.
[23] Liu Wei, Anguelov D, Erhan D, et al.SSD: single shot MultiBox detector[M]//Computer Vision—ECCV 2016. Cham: Springer International Publishing, 2016.
[24] Qassim H, Verma A, Feinzimer D.Compressed residual-VGG16 CNN model for big data places image recognition[C]//2018 IEEE 8th Annual Com- puting and Communication Workshop and Conference, Las Vegas, NV, USA, 2018: 169-175.
[25] 曾升, 耿国华, 邹林波, 等. 第一人称视角地形轮廓草图的真实空间重建[J]. 光学精密工程, 2020, 28(8): 1861-1871.
Zeng Sheng, Geng Guohua, Zou Linbo, et al.Real spatial terrain reconstruction of first person point- of-view sketches[J]. Optics and Precision Engineering, 2020, 28(8): 1861-1871.
[26] 李峰. 智能监控中对运动目标的检测和测距技术研究[D]. 合肥: 合肥工业大学, 2013. |
|
|
|
2023, Vol. 38 
