Aircraft Target Detection in Remote Sensing Images Based on Improved Faster R-CNN

This paper proposes an improved Faster R-CNN algorithm to improve the accuracy of aircraft detection in remote sensing images. The detection accuracy of existing algorithms is low in the case of complex background, dense target and poor image quality, so a more efficient and accurate algorithm is needed to solve this problem. The improved algorithm first improves the feature extraction network by using an efficient attention fusion module, thus improving the accuracy and efficiency of feature extraction. The Area Suggestion Network (RPN) is then used to generate candidate regions to improve the positioning accuracy of aircraft targets. Finally, the classification regression network of Faster R-CNN is used to obtain the detection results of aircraft targets to further improve the accuracy. In this paper, the UCAS-AOD dataset is used for experimental evaluation and comparison with other similar algorithms. The experimental results show that the algorithm achieves an average detection accuracy of 92.56% and an FPS of 41.97 with a few additional parameters, which proves that the algorithm has excellent performance and generalization ability in aircraft target detection in remote sensing images.