UAV detection under low-light conditions poses a significant challenge in computer vision and UAV surveillance, as factors like insufficient illumination, noise interference, and small target sizes cause traditional detection methods to exhibit significant performance degradation. This paper proposes a deep learning-based UAV object detection algorithm for low-light environments, integrating image enhancement and object detection to improve accuracy and robustness. First, a low-light image enhancement network (DDNet) is applied for preprocessing. The detection network employs the YOLOv11 algorithm, enhanced with a Synergistic Channel Spatial Pixel (SCSP) module to strengthen perception capability for challenging features like low contrast and shape variations. Finally, a Local Feature Embedded Global Feature Extraction Module (LEGM) is introduced to capture both local and global features. Experimental results demonstrate that the proposed method achieves 3.2% and 1.3% improvements in mAP@0.5 and mAP@0.5:0.95, respectively, on the VisDrone2019 (dark) datasets. This study shows promising application prospects for UAV object detection in low-light environments.