The paper presents EDGS-YOLOv8, an improved lightweight UAV detection model based on YOLOv8. The model is designed to enhance the accuracy, real-time performance, and efficiency of drone detection, particularly for small targets. Key improvements include: 1. **Network Neck Structure**: The neck structure is enhanced by introducing the C3Ghost module and ghost convolution, which reduces computational complexity and improves feature extraction. 2. **Efficient Multi-Scale Attention (EMA)**: The EMA mechanism is integrated into the neck to enhance the model's ability to capture multi-scale spatial details and improve detection accuracy. 3. **DDetect Module**: The DDetect module, incorporating deformable convolution (DCNv2), is designed to improve the model's robustness and accuracy in detecting UAV features, especially in complex scenes. 4. **Model Evaluation**: The proposed model is evaluated using the DUT Anti-UAV dataset, achieving an AP value of 0.971, 3.1% higher than YOLOv8n's mAP, while maintaining a model size of only 4.23 MB. The model also shows superior performance on the Det-Fly dataset and the VisDrone2019 dataset for small target detection. The results demonstrate that EDGS-YOLOv8 effectively balances accuracy, compactness, and real-time performance, making it suitable for embedded devices and mobile terminals. The model's improvements in accuracy, recall, and mAP50/mAP95, along with its smaller model size and higher FPS, highlight its advantages over other state-of-the-art models.The paper presents EDGS-YOLOv8, an improved lightweight UAV detection model based on YOLOv8. The model is designed to enhance the accuracy, real-time performance, and efficiency of drone detection, particularly for small targets. Key improvements include: 1. **Network Neck Structure**: The neck structure is enhanced by introducing the C3Ghost module and ghost convolution, which reduces computational complexity and improves feature extraction. 2. **Efficient Multi-Scale Attention (EMA)**: The EMA mechanism is integrated into the neck to enhance the model's ability to capture multi-scale spatial details and improve detection accuracy. 3. **DDetect Module**: The DDetect module, incorporating deformable convolution (DCNv2), is designed to improve the model's robustness and accuracy in detecting UAV features, especially in complex scenes. 4. **Model Evaluation**: The proposed model is evaluated using the DUT Anti-UAV dataset, achieving an AP value of 0.971, 3.1% higher than YOLOv8n's mAP, while maintaining a model size of only 4.23 MB. The model also shows superior performance on the Det-Fly dataset and the VisDrone2019 dataset for small target detection. The results demonstrate that EDGS-YOLOv8 effectively balances accuracy, compactness, and real-time performance, making it suitable for embedded devices and mobile terminals. The model's improvements in accuracy, recall, and mAP50/mAP95, along with its smaller model size and higher FPS, highlight its advantages over other state-of-the-art models.