Detection of Bird Nests on Transmission Towers in Aerial Images Based on Improved YOLOv5s
To further improve the accuracy of bird nest model detection on transmission towers in aerial images without significantly increasing the model size and to make detection more suitable for edge-end applications, the lightweight model YOLOv5s is improved in this paper. First, the original backbone ne...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-02-01
|
| Series: | Machines |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-1702/11/2/257 |
| _version_ | 1797619734491955200 |
|---|---|
| author | Gujing Han Ruijie Wang Qiwei Yuan Saidian Li Liu Zhao Min He Shiqi Yang Liang Qin |
| author_facet | Gujing Han Ruijie Wang Qiwei Yuan Saidian Li Liu Zhao Min He Shiqi Yang Liang Qin |
| author_sort | Gujing Han |
| collection | DOAJ |
| description | To further improve the accuracy of bird nest model detection on transmission towers in aerial images without significantly increasing the model size and to make detection more suitable for edge-end applications, the lightweight model YOLOv5s is improved in this paper. First, the original backbone network is reconfigured using the OSA (One-Shot Aggregation) module in the VOVNet and the CBAM (Convolution Block Attention Module) is embedded into the feature extraction network, which improves the accuracy of the model for small target recognition. Then, the atrous rates and the number of atrous convolutions of the ASPP (Atrous Spatial Pyramid Pooling) module are reduced to effectively decrease the parameters of the ASPP. The ASPP is then embedded into the feature fusion network to enhance the detection of the targets in complex backgrounds, improving the model accuracy. The experiments show that the mAP (mean-Average Precision) of the fusion-improved YOLOv5s model improves from 91.84% to 95.18%, with only a 27.4% increase in model size. Finally, the improved YOLOv5s model is deployed into the Jeston Xavier NX, resulting in a model that runs well and has a substantial increase in accuracy and a speed of 10.2 FPS, which is only 0.7 FPS slower than the original YOLOv5s model. |
| first_indexed | 2024-03-11T08:31:05Z |
| format | Article |
| id | doaj.art-081e3e896fe24073afd2b49dc8c95c49 |
| institution | Directory Open Access Journal |
| issn | 2075-1702 |
| language | English |
| last_indexed | 2024-03-11T08:31:05Z |
| publishDate | 2023-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Machines |
| spelling | doaj.art-081e3e896fe24073afd2b49dc8c95c492023-11-16T21:46:04ZengMDPI AGMachines2075-17022023-02-0111225710.3390/machines11020257Detection of Bird Nests on Transmission Towers in Aerial Images Based on Improved YOLOv5sGujing Han0Ruijie Wang1Qiwei Yuan2Saidian Li3Liu Zhao4Min He5Shiqi Yang6Liang Qin7Department of Electronic and Electrical Engineering, Wuhan Textile University, Wuhan 430200, ChinaDepartment of Electronic and Electrical Engineering, Wuhan Textile University, Wuhan 430200, ChinaDepartment of Electronic and Electrical Engineering, Wuhan Textile University, Wuhan 430200, ChinaDepartment of Electronic and Electrical Engineering, Wuhan Textile University, Wuhan 430200, ChinaDepartment of Electronic and Electrical Engineering, Wuhan Textile University, Wuhan 430200, ChinaSchool of Electrical and Automation, Wuhan University, Wuhan 430072, ChinaSchool of Electrical and Automation, Wuhan University, Wuhan 430072, ChinaSchool of Electrical and Automation, Wuhan University, Wuhan 430072, ChinaTo further improve the accuracy of bird nest model detection on transmission towers in aerial images without significantly increasing the model size and to make detection more suitable for edge-end applications, the lightweight model YOLOv5s is improved in this paper. First, the original backbone network is reconfigured using the OSA (One-Shot Aggregation) module in the VOVNet and the CBAM (Convolution Block Attention Module) is embedded into the feature extraction network, which improves the accuracy of the model for small target recognition. Then, the atrous rates and the number of atrous convolutions of the ASPP (Atrous Spatial Pyramid Pooling) module are reduced to effectively decrease the parameters of the ASPP. The ASPP is then embedded into the feature fusion network to enhance the detection of the targets in complex backgrounds, improving the model accuracy. The experiments show that the mAP (mean-Average Precision) of the fusion-improved YOLOv5s model improves from 91.84% to 95.18%, with only a 27.4% increase in model size. Finally, the improved YOLOv5s model is deployed into the Jeston Xavier NX, resulting in a model that runs well and has a substantial increase in accuracy and a speed of 10.2 FPS, which is only 0.7 FPS slower than the original YOLOv5s model.https://www.mdpi.com/2075-1702/11/2/257aerial imagesbird nest detectionYOLOv5smodel deployment |
| spellingShingle | Gujing Han Ruijie Wang Qiwei Yuan Saidian Li Liu Zhao Min He Shiqi Yang Liang Qin Detection of Bird Nests on Transmission Towers in Aerial Images Based on Improved YOLOv5s Machines aerial images bird nest detection YOLOv5s model deployment |
| title | Detection of Bird Nests on Transmission Towers in Aerial Images Based on Improved YOLOv5s |
| title_full | Detection of Bird Nests on Transmission Towers in Aerial Images Based on Improved YOLOv5s |
| title_fullStr | Detection of Bird Nests on Transmission Towers in Aerial Images Based on Improved YOLOv5s |
| title_full_unstemmed | Detection of Bird Nests on Transmission Towers in Aerial Images Based on Improved YOLOv5s |
| title_short | Detection of Bird Nests on Transmission Towers in Aerial Images Based on Improved YOLOv5s |
| title_sort | detection of bird nests on transmission towers in aerial images based on improved yolov5s |
| topic | aerial images bird nest detection YOLOv5s model deployment |
| url | https://www.mdpi.com/2075-1702/11/2/257 |
| work_keys_str_mv | AT gujinghan detectionofbirdnestsontransmissiontowersinaerialimagesbasedonimprovedyolov5s AT ruijiewang detectionofbirdnestsontransmissiontowersinaerialimagesbasedonimprovedyolov5s AT qiweiyuan detectionofbirdnestsontransmissiontowersinaerialimagesbasedonimprovedyolov5s AT saidianli detectionofbirdnestsontransmissiontowersinaerialimagesbasedonimprovedyolov5s AT liuzhao detectionofbirdnestsontransmissiontowersinaerialimagesbasedonimprovedyolov5s AT minhe detectionofbirdnestsontransmissiontowersinaerialimagesbasedonimprovedyolov5s AT shiqiyang detectionofbirdnestsontransmissiontowersinaerialimagesbasedonimprovedyolov5s AT liangqin detectionofbirdnestsontransmissiontowersinaerialimagesbasedonimprovedyolov5s |