This paper presents an improved YOLOv8 algorithm for detecting insulator defects in drone aerial photography. The authors introduce a Multi-scale Large Kernel Attention (MLKA) module to enhance the model's focus on features of different scales and low-level feature maps. They also employ lightweight GSCConv convolution and construct the GSC_C2f module to simplify the computational process and reduce memory burden, thereby improving the performance of insulator defect detection. Additionally, an improved loss function using SiOU is adopted to optimize the model's detection performance and enhance its feature extraction capability for insulator defects. Experimental results demonstrate that the improved model achieves an mAP of 99.22% and an FPS of 55.73, outperforming the original YOLOv8s and YOLOv5s models by 2.18% and 2.91%, respectively. The model size is only 30.18MB, making it suitable for real-time operation and high accuracy. The paper also includes a detailed analysis of the model's performance, including comparative experiments with other advanced models and robustness tests under challenging environmental conditions.This paper presents an improved YOLOv8 algorithm for detecting insulator defects in drone aerial photography. The authors introduce a Multi-scale Large Kernel Attention (MLKA) module to enhance the model's focus on features of different scales and low-level feature maps. They also employ lightweight GSCConv convolution and construct the GSC_C2f module to simplify the computational process and reduce memory burden, thereby improving the performance of insulator defect detection. Additionally, an improved loss function using SiOU is adopted to optimize the model's detection performance and enhance its feature extraction capability for insulator defects. Experimental results demonstrate that the improved model achieves an mAP of 99.22% and an FPS of 55.73, outperforming the original YOLOv8s and YOLOv5s models by 2.18% and 2.91%, respectively. The model size is only 30.18MB, making it suitable for real-time operation and high accuracy. The paper also includes a detailed analysis of the model's performance, including comparative experiments with other advanced models and robustness tests under challenging environmental conditions.