Remote sensing small target detection method based on SPL-RTDETR

Remote sensing satellite images often face challenges such as excessively small detection targets and difficulties in feature extraction, leading to low accuracy of existing target detection models in remote sensing small target detection. To address these issues, this paper proposes an SPL-RTDETR-based method for small target detection in remote sensing. First, the SAHI method is employed to slice the original remote sensing images, mitigating issues of extremely small target sizes and dense distributions. Second, a PPA module is introduced to preserve and enhance small target features, ensuring that critical information of small targets is retained after multiple downsampling steps. Finally, an LSTF layer is added to incorporate shallow image features for subsequent fusion processing, improving the model's learning capability for small targets. Experimental results demonstrate that the proposed method achieves higher detection accuracy than existing target detection models, exhibiting superior precision and robustness in remote sensing small target detection.