18 September 2023 in IEEE Access
The insulator defect targets in UAV (Unmanned Aerial Vehicle) images are often small and set against complex backgrounds. Consequently, traditional object detection algorithms commonly struggle to identify these minor defects. To enhance precision and recall in detecting insulator defects, a novel model, GC-YOLO (ghost convolution and centralized feature pyramid -You Only Look Once), based on YOLOv5s, has been introduced. GC-YOLO incorporates the Ghost convolution module in the backbone network, reducing feature redundancy and improving the inference speed of the feature extraction network. Moreover, an attention mechanism based on Coordinate Attention (CA) is integrated at the terminal of the backbone network, aimed at emphasizing the extraction of crucial information from target features. The Explicit Visual Center Block (EVCBlock) module from Centralized Feature Pyramid Network (CFPNet) is introduced in the neck layer to effectively fuse multi-scale features and enhance the feature map’s characterization capability. Furthermore, in order to enhance the precision in detecting small-sized defects, a small object detection head is also added to the detection layer based on CFPNet. Experimental results demonstrate that GC-YOLO achieves a recall of 89.7% and mAP@0.5 of 94.2%, surpassing YOLOv5s by 7% and 6.5%, respectively. The proposed algorithm exhibits superior detection precision in complex scenes, providing a theoretical basis for intelligent and mechanized railway monitoring systems.
2 October 2023 in IEEE Access
Ensuring the promptly removal of foreign objects from transmission lines is crucial for electricity safety. However, the existing object detection algorithms exhibit low precision and recall due to factors such as uncertainty in the type of foreign objects, imbalance of positive and negative samples, and the complexity of aerial photography backgrounds. Therefore, in this paper, an algorithm called DF-YOLO (Deformable Faster-You Only Look Once) for transmission line foreign object detection is proposed to resolve these problems. The algorithm is based on YOLOv7-Tiny and tailored to the dataset’s characteristics to achieve high precision along with excellent recall. The Focal-DIoU loss function is utilized to balance positive and negative sample proportions during training. Additionally, the algorithm incorporates the deformable convolution (DCN) module and the SimAM attention mechanism to enhance model performance, particularly in terms of foreign object recall and detection accuracy. Moreover, we optimized the network inference speed with the improved SPPCSPC_S-F module. Experimental results demonstrate that the improved DF-YOLO network achieves a 2.04% increase in mAP@.5 compared to the original YOLOv7-Tiny network. The recall rate also improves from 89% to 91.51%. Additionally, the inference speed of the network rises from 130 to 140 FPS, which enhances detection effectiveness and reduces the frequency of transmission line leaks triggered by foreign object incursion.