2019/11/20 by Saeed Afshar, Andrew P Nicholson, Afshar, Saeed +5 · 1 citation
Engineering · Neuroscience · #Advanced Memory and Neural Computing #CCD and CMOS Imaging Sensors #Computer Vision and Pattern Recognition (cs.CV) #EEG and Brain-Computer Interfaces #FOS: Computer and information sciences
paper · pdf · doi:10.48550/arxiv.1911.08730
openalex publication_date 2019/11/20 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
In this work, we present optical space imaging using an unconventional yet\npromising class of imaging devices known as neuromorphic event-based sensors.\nThese devices, which are modeled on the human retina, do not operate with\nframes, but rather generate asynchronous streams of events in response to\nchanges in log-illumination at each pixel. These devices are therefore\nextremely fast, do not have fixed exposure times, allow for imaging whilst the\ndevice is moving and enable low power space imaging during daytime as well as\nnight without modification of the sensors. Recorded at multiple remote sites,\nwe present the first event-based space imaging dataset including recordings\nfrom multiple event-based sensors from multiple providers, greatly lowering the\nbarrier to entry for other researchers given the scarcity of such sensors and\nthe expertise required to operate them. The dataset contains 236 separate\nrecordings and 572 labeled resident space objects. The event-based imaging\nparadigm presents unique opportunities and challenges motivating the\ndevelopment of specialized event-based algorithms that can perform tasks such\nas detection and tracking in an event-based manner. Here we examine a range of\nsuch event-based algorithms for detection and tracking. The presented methods\nare designed specifically for space situational awareness applications and are\nevaluated in terms of accuracy and speed and suitability for implementation in\nneuromorphic hardware on remote or space-based imaging platforms.\n