Improving Neural Network Detection Accuracy of Electric Power Bushings in Infrared Images by Hough Transform
To improve the neural network detection accuracy of the electric power bushings in infrared images, a modified algorithm based on the You Only Look Once version 2 (YOLOv2) network is proposed to achieve better recognition results. Specifically, YOLOv2 corresponds to a convolutional neural network (C...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-05-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/20/10/2931 |
| _version_ | 1797567342685716480 |
|---|---|
| author | Hongshan Zhao Zeyan Zhang |
| author_facet | Hongshan Zhao Zeyan Zhang |
| author_sort | Hongshan Zhao |
| collection | DOAJ |
| description | To improve the neural network detection accuracy of the electric power bushings in infrared images, a modified algorithm based on the You Only Look Once version 2 (YOLOv2) network is proposed to achieve better recognition results. Specifically, YOLOv2 corresponds to a convolutional neural network (CNN), although its rotation invariance is poor, and some bounding boxes (BBs) exhibit certain deviations. To solve this problem, the standard Hough transform and image rotation are utilized to determine the optimal recognition angle for target detection, such that an optimal recognition effect of YOLOv2 on inclined objects (for example, bushing) is achieved. With respect to the problem that the BB is biased, the shape feature of the bushing is extracted by the Gap statistic algorithm, based on K-means clustering; thereafter, the sliding window (SW) is utilized to determine the optimal recognition area. Experimental verification indicates that the proposed rotating image method can improve the recognition effect, and the SW can further modify the BB. The accuracy of target detection increases to 97.33%, and the recall increases to 95%. |
| first_indexed | 2024-03-10T19:40:22Z |
| format | Article |
| id | doaj.art-19ff0b4e8a594aaf957a2c77158ef77b |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T19:40:22Z |
| publishDate | 2020-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-19ff0b4e8a594aaf957a2c77158ef77b2023-11-20T01:17:20ZengMDPI AGSensors1424-82202020-05-012010293110.3390/s20102931Improving Neural Network Detection Accuracy of Electric Power Bushings in Infrared Images by Hough TransformHongshan Zhao0Zeyan Zhang1School of Electrical & Electronic Engineering, North China Electric Power University, Baoding 071003, ChinaSchool of Electrical & Electronic Engineering, North China Electric Power University, Baoding 071003, ChinaTo improve the neural network detection accuracy of the electric power bushings in infrared images, a modified algorithm based on the You Only Look Once version 2 (YOLOv2) network is proposed to achieve better recognition results. Specifically, YOLOv2 corresponds to a convolutional neural network (CNN), although its rotation invariance is poor, and some bounding boxes (BBs) exhibit certain deviations. To solve this problem, the standard Hough transform and image rotation are utilized to determine the optimal recognition angle for target detection, such that an optimal recognition effect of YOLOv2 on inclined objects (for example, bushing) is achieved. With respect to the problem that the BB is biased, the shape feature of the bushing is extracted by the Gap statistic algorithm, based on K-means clustering; thereafter, the sliding window (SW) is utilized to determine the optimal recognition area. Experimental verification indicates that the proposed rotating image method can improve the recognition effect, and the SW can further modify the BB. The accuracy of target detection increases to 97.33%, and the recall increases to 95%.https://www.mdpi.com/1424-8220/20/10/2931infrared imagepower apparatus bushingstandard hough transformtarget detectionYOLOv2CNN |
| spellingShingle | Hongshan Zhao Zeyan Zhang Improving Neural Network Detection Accuracy of Electric Power Bushings in Infrared Images by Hough Transform Sensors infrared image power apparatus bushing standard hough transform target detection YOLOv2 CNN |
| title | Improving Neural Network Detection Accuracy of Electric Power Bushings in Infrared Images by Hough Transform |
| title_full | Improving Neural Network Detection Accuracy of Electric Power Bushings in Infrared Images by Hough Transform |
| title_fullStr | Improving Neural Network Detection Accuracy of Electric Power Bushings in Infrared Images by Hough Transform |
| title_full_unstemmed | Improving Neural Network Detection Accuracy of Electric Power Bushings in Infrared Images by Hough Transform |
| title_short | Improving Neural Network Detection Accuracy of Electric Power Bushings in Infrared Images by Hough Transform |
| title_sort | improving neural network detection accuracy of electric power bushings in infrared images by hough transform |
| topic | infrared image power apparatus bushing standard hough transform target detection YOLOv2 CNN |
| url | https://www.mdpi.com/1424-8220/20/10/2931 |
| work_keys_str_mv | AT hongshanzhao improvingneuralnetworkdetectionaccuracyofelectricpowerbushingsininfraredimagesbyhoughtransform AT zeyanzhang improvingneuralnetworkdetectionaccuracyofelectricpowerbushingsininfraredimagesbyhoughtransform |