Research on Insulator Defect Detection by UAV Based on Deep Learning

To address the limitations of existing insulator detection algorithms in accurately identifying insulators and detecting multiple defects, a novel approach is proposed based on the improved YOLOv7 algorithm. The method incorporates several key enhancements. Firstly, the MPDlo U loss function is introduced to simplify the calculation process. Secondly, the lightweight CBAM attention mechanism is integrated into the Hand part to enhance global perception of small target defects. Lastly, the SPPFCSPC pyramid pooling is incorporated to improve the recognition and detection capabilities of feature maps. Experimental results demonstrate that the improved algorithm achieves an mAP of 95%, an accuracy rate of 94.5%, and a recall rate of 92.6%.