Automated Military Vehicle detection from Aerial Imagery

Abstract

Detecting military vehicles and distinguishing them out from non-military vehicles is a significant challenge in the defence sector. Detection of military vehicle could help to identify enemy’s move and hence, build early precau tionary measures. Recently, many deep learning based techniques have been proposed for vehicle detection purpose. However, they are developed using datasets that are not useful if military specific vehicle training and detection is required. Hyper-parameters in those techniques are not tuned to enter tain low-altitude aerial imagery. The thesis aimed at developing or adapting state-of-the-art deep learning frameworks to detect particularly military ve hicle along with other standard non-military vehicles. The major bottleneck in the application of deep learning frameworks to detect military vehicles is the lack of available datasets. In this context, we prepared a dataset of low-altitude aerial images that comprises of real data (taken from military shows videos) and toy data (taken from YouTube videos). Our dataset is categorized into two main types i.e. military and non-military vehicles. We employed state-of-the-art object detection algorithms to distinguish military and non-military vehicles. Specifically, the three deep architectures used for this purpose include faster region-based convolutional neural networks, recurrent fully convolutional neural networks, and single shot multibox de tector (SSD). We also did comparative analysis of architectures by increasing training data and observing it’s impact on results. The experimental results show that the training of deep architectures using the customized/prepared dataset allows to recognize seven types of military and four types of non military vehicles. It can handle complex scenarios by differentiating vehicle from it’s surroundings objects. We report the mean average precision (MAP) obtained using the three adopted architectures with SSD giving the highest mean average precision of around 77.67% with 800,000 iterations (3 epochs).