No abstract

Distributed unattended ground sensor (UGS) networks used in battlefield surveillance and monitoring missions, have proven to be valuable in providing a tactical information advantage required for command and control, intelligence, surveillance and reconnaissance planning. Operational effectiveness for surveillance missions can be enhanced further through network centric capability (NCC), where distributed UGS networks have the ability to perform surveillance operations autonomously. NCC operation can be enhanced through netwoked UGSs having the ability to evaluate their awareness of the current joint surveillance environment, in order to provide the necessary adaption to dynamic changes. NCC can also provide an advantage for UGS networks to selfmanage their limited operational resources efficiently, according to mission objective priority. In this article, we present a cross-layer approach and highlight techniques, which have potential to enable NCC operation within a mission-orientated UGS surveillance setting.

Taking inspiration from the structure and behaviour of the human visual system and using the Transposed Convolution and Saliency Mapping methods of Convolutional Neural Networks (CNN), a spiking event-based image segmentation algorithm, SpikeSEG is proposed. The approach makes use of both spike-based imaging and spike-based processing, where the images are either standard images converted to spiking images or they are generated directly from a neuromorphic event driven sensor, and then processed using a spiking fully convolutional neural network. The spiking segmentation method uses the spike activations through time within the network to trace back any outputs from saliency maps, to the exact pixel location. This not only gives exact pixel locations for spiking segmentation, but with low latency and computational overhead. SpikeSEG is the first spiking event-based segmentation network and over three experiment test achieves promising results with 96% accuracy overall and a 74% mean intersection over union for the segmentation, all within an event by event-based framework.

No abstract

No abstract