“In the field of fine-grained recognition of remote sensing aircraft, experts have proposed the YOLOv5s algorithm that integrates bounding box Gaussian modeling and feature aggregation distribution, effectively improving the accuracy of small object detection. The experimental results showed that the model accuracy reached 99.10% and 95.36% respectively, with the best detection accuracy.”