Remote Sensing Object Detection Method Based on Attention Mechanism and Multi-scale Feature Fusion

Aiming at the low target detection accuracy due to the complex imaging background, small size, large number and dense arrangement of targets in optical remote sensing images, this paper was proposed an improved Faster R-CNN algorithm. Firstly, the attention mechanism module is reasonably embedded in the backbone network to emphasize the target information and suppress the background information; Secondly, a feature fusion method is designed to fuse the information of each feature layer in the backbone network to improve the small target detection ability; In addition, the loss function and pooling function are improved and the Anchor parameters are optimized. It can be seen from the experimental data that the detection accuracy of the method proposed in this paper is significantly improved compared with the original algorithm, and the detection accuracy is more dominant than other mainstream algorithms.