To improve the peak signal-to-noise ratio and enhance the detailed features of drone aerial images, this study proposes a new method for enhancing the detailed features of drone aerial images using graph neural networks. Firstly, construct a Graph Neural Network (GNN) model to better handle complex scene relationships in images, thereby extracting features from drone aerial images; Then, through extensive feature learning, automatically adjust the enhancement level of different regions or features; Finally, reconstruct the image by intelligently fusing the enhanced features with the original image features, ensuring that the clarity and color information of the original image are maximally preserved while enhancing details. The test results show that after applying this method, the peak signal-to-noise ratio of unmanned aerial vehicle (UAV) aerial images is significantly higher than that of conventional methods, indicating that this method can greatly enhance image features and optimize image quality.