Abstract: Aircraft detection is one of the hottest issues in the field of remote sensing image analysis. The current detection methods of remote sensing image exist many problems, such as complex detection procedure, low accuracy in complex background and dense aircraft area. To solve these problems, an end-to-end aircraft detection method named MRNSSD (Multiscale Residual Network Single Shot Detector) is proposed. In this framework, a residual network is used to extract features for its powerful ability in feature extraction, then an extra sub-network consisting of several feature layers is appended to detect and locate aircrafts. In order to locate aircrafts of various scales more accurately, a series of aspect ratios of default boxes are set to better match aircraft shapes and combine predictions deduced from feature maps of different layers. The method is more brief and efficient than methods that require object proposals, because it eliminates proposal generation completely and encapsulates all computation in a single network. Experiments demonstrate that this approach achieves better performance in many complex scenes.

Key words: aircraft detection, remote sensing image, deep convolutional network