Aircraft detection using deep convolutional neural networkbased semantic segmentation

Abstract

This paper proposes a vision-based aircraft detection method using deep convolutional neural network-based semantic segmentation. SegNet architecture is used as deep learning model. The proposed method detects aircrafts even with a complex background and can be applied to various environments without modification of the algorithm. Several videos were tested to verify the performance of the proposed method, and the results are shown and summarized for each case. The results show robustness against scale, light intensity, and viewpoint for flight detection.