“In the field of deep learning, experts have proposed a remote sensing image data augmentation algorithm based on diffusion models, which solves the problems of sample scarcity and imbalance, significantly improves the accuracy of model detection and cross dataset adaptability, and provides technical support for small sample remote sensing archaeological target recognition.”