Remote Sensing Image Target Detection Algorithm Based on Multi-dimension Attention
Pengfei Shi, Rong Fan, Wei Yan, Xuan Wu Zhou, Xinyang Chen · 2021 7th International Conference on Computer and Communications (ICCC) · 2021
Aiming at the problem of missing detection caused by the densely arranged targets in remote sensing images, a target detection algorithm based on multi-dimension attention is proposed. The proposed algorithm uses the RetinaNet target detector as the main framework. Firstly, the backbone network is used to extract feature maps of different scales. Secondly, the Feature Pyramid Net (FPN) is used to realize the fusion of feature maps of different scales. Then, the feature map through the attention module is used to suppress the boundary noise and highlight the target features. Finally, the class and box subnets are used to classify and locate the target. The experimental results show that, compared with the RetinaNet algorithm, the mean Average Precision (mAP) of the improved algorithm on small-DOTA is improved by 5.55%.