Automatic infrared image recognition method for substation equipment based on a deep self-attention network and multi-factor similarity calculation
Infrared image recognition plays an important role in the inspection of power equipment. Existing technologies dedicated to this purpose often require manually selected features, which are not transferable and interpretable, and have limited training data. To address these limitations, this paper pr...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
KeAi Communications Co., Ltd.
2022-08-01
|
| Series: | Global Energy Interconnection |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2096511722000792 |
| _version_ | 1828178284815319040 |
|---|---|
| author | Yaocheng Li Yongpeng Xu Mingkai Xu Siyuan Wang Zhicheng Xie Zhe Li Xiuchen Jiang |
| author_facet | Yaocheng Li Yongpeng Xu Mingkai Xu Siyuan Wang Zhicheng Xie Zhe Li Xiuchen Jiang |
| author_sort | Yaocheng Li |
| collection | DOAJ |
| description | Infrared image recognition plays an important role in the inspection of power equipment. Existing technologies dedicated to this purpose often require manually selected features, which are not transferable and interpretable, and have limited training data. To address these limitations, this paper proposes an automatic infrared image recognition framework, which includes an object recognition module based on a deep self-attention network and a temperature distribution identification module based on a multi-factor similarity calculation. First, the features of an input image are extracted and embedded using a multi-head attention encoding–decoding mechanism. Thereafter, the embedded features are used to predict the equipment component category and location. In the located area, preliminary segmentation is performed. Finally, similar areas are gradually merged, and the temperature distribution of the equipment is obtained to identify a fault. Our experiments indicate that the proposed method demonstrates significantly improved accuracy compared with other related methods and, hence, provides a good reference for the automation of power equipment inspection. |
| first_indexed | 2024-04-12T05:13:55Z |
| format | Article |
| id | doaj.art-71f997f9cc084ba885e5640a71ea31e8 |
| institution | Directory Open Access Journal |
| issn | 2096-5117 |
| language | English |
| last_indexed | 2024-04-12T05:13:55Z |
| publishDate | 2022-08-01 |
| publisher | KeAi Communications Co., Ltd. |
| record_format | Article |
| series | Global Energy Interconnection |
| spelling | doaj.art-71f997f9cc084ba885e5640a71ea31e82022-12-22T03:46:41ZengKeAi Communications Co., Ltd.Global Energy Interconnection2096-51172022-08-0154397408Automatic infrared image recognition method for substation equipment based on a deep self-attention network and multi-factor similarity calculationYaocheng Li0Yongpeng Xu1Mingkai Xu2Siyuan Wang3Zhicheng Xie4Zhe Li5Xiuchen Jiang6School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, 200240 Shanghai, PR ChinaSchool of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, 200240 Shanghai, PR ChinaState Grid Shandong Electric Power Company Jinan Power Supply Company, 250001 Shandong, PR ChinaState Grid Shandong Electric Power Company Jinan Power Supply Company, 250001 Shandong, PR ChinaMaintenance & Test Center of EHV power Transmission Company, China Southern Power Grid, 510000 Guangdong, PR ChinaSchool of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, 200240 Shanghai, PR ChinaSchool of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, 200240 Shanghai, PR ChinaInfrared image recognition plays an important role in the inspection of power equipment. Existing technologies dedicated to this purpose often require manually selected features, which are not transferable and interpretable, and have limited training data. To address these limitations, this paper proposes an automatic infrared image recognition framework, which includes an object recognition module based on a deep self-attention network and a temperature distribution identification module based on a multi-factor similarity calculation. First, the features of an input image are extracted and embedded using a multi-head attention encoding–decoding mechanism. Thereafter, the embedded features are used to predict the equipment component category and location. In the located area, preliminary segmentation is performed. Finally, similar areas are gradually merged, and the temperature distribution of the equipment is obtained to identify a fault. Our experiments indicate that the proposed method demonstrates significantly improved accuracy compared with other related methods and, hence, provides a good reference for the automation of power equipment inspection.http://www.sciencedirect.com/science/article/pii/S2096511722000792Substation equipmentInfrared image intelligent recognitionDeep self-attention networkMulti-factor similarity calculation |
| spellingShingle | Yaocheng Li Yongpeng Xu Mingkai Xu Siyuan Wang Zhicheng Xie Zhe Li Xiuchen Jiang Automatic infrared image recognition method for substation equipment based on a deep self-attention network and multi-factor similarity calculation Global Energy Interconnection Substation equipment Infrared image intelligent recognition Deep self-attention network Multi-factor similarity calculation |
| title | Automatic infrared image recognition method for substation equipment based on a deep self-attention network and multi-factor similarity calculation |
| title_full | Automatic infrared image recognition method for substation equipment based on a deep self-attention network and multi-factor similarity calculation |
| title_fullStr | Automatic infrared image recognition method for substation equipment based on a deep self-attention network and multi-factor similarity calculation |
| title_full_unstemmed | Automatic infrared image recognition method for substation equipment based on a deep self-attention network and multi-factor similarity calculation |
| title_short | Automatic infrared image recognition method for substation equipment based on a deep self-attention network and multi-factor similarity calculation |
| title_sort | automatic infrared image recognition method for substation equipment based on a deep self attention network and multi factor similarity calculation |
| topic | Substation equipment Infrared image intelligent recognition Deep self-attention network Multi-factor similarity calculation |
| url | http://www.sciencedirect.com/science/article/pii/S2096511722000792 |
| work_keys_str_mv | AT yaochengli automaticinfraredimagerecognitionmethodforsubstationequipmentbasedonadeepselfattentionnetworkandmultifactorsimilaritycalculation AT yongpengxu automaticinfraredimagerecognitionmethodforsubstationequipmentbasedonadeepselfattentionnetworkandmultifactorsimilaritycalculation AT mingkaixu automaticinfraredimagerecognitionmethodforsubstationequipmentbasedonadeepselfattentionnetworkandmultifactorsimilaritycalculation AT siyuanwang automaticinfraredimagerecognitionmethodforsubstationequipmentbasedonadeepselfattentionnetworkandmultifactorsimilaritycalculation AT zhichengxie automaticinfraredimagerecognitionmethodforsubstationequipmentbasedonadeepselfattentionnetworkandmultifactorsimilaritycalculation AT zheli automaticinfraredimagerecognitionmethodforsubstationequipmentbasedonadeepselfattentionnetworkandmultifactorsimilaritycalculation AT xiuchenjiang automaticinfraredimagerecognitionmethodforsubstationequipmentbasedonadeepselfattentionnetworkandmultifactorsimilaritycalculation |