Remote Sensing Image Aircraft Detection Technology Based on Deep Learning
Wanjun Wei, Jiuwen Zhang · 2019
How to use high-resolution remote sensing images to quickly and accurately obtain high-value target information such as aircraft and ships has become a hot topic in the research of automatic target detection. This paper uses Google Earth image to make aircraft data set, then detect the aircraft targets respectively based on the two deep learning models YOLOv3 and Faster_R_CNN. At the same time, in order to significantly improve the aircraft detection accuracy, this paper proposes a shadow processing algorithm with double threshold random sampling to conduct data preprocessing for aircraft targets in remote sensing images. The experimental results show that both deep learning models can effectively detect aircraft targets and have great application potential in automatic detection of remote sensing image targets. The shadow processing algorithm can effectively eliminate the projected shadow of the aircraft, restore the texture characteristics of the shaded area, greatly improve the overall quality of the remote sensing image, and lay a good data foundation for subsequent aircraft detection.