Accurate detection of cotton leaf diseases in complex natural environments is essential for effectively minimizing the impact of diseases on cotton yield and quality.To address this，this study proposes a cotton leaf disease detection model based on an improved YOLOv8n.First，the EMA attention mechanism is integrated into the backbone network of YOLOv8n.Simultaneously，the deformable convolutional module，Deformable ConvNets v2，is incorporated into the C2f module within the backbone network to expand the sensory field and strengthen feature extraction capabilities.Additionally，the CIoU loss function is replaced with the WIoU bounding box regression loss，which includes a dynamic focusing mechanism to accelerate model convergence and further improve performance.The experimental results showed that the improved EDW-YOLOv8 model achieves increases of 4.3%，7.5% and 4.6% in accuracy，recall，and average precision，respectively，compared with the original YOLOv8n.These results show that the proposed model，with good generalization ability，can accurately and efficiently detect cotton leaf disease targets in images.