High-Speed Image Processing in Heavily Degraded Environments Using a Machine Learning Classifier

Driving ground vehicles in heavily degraded visual environments brings significant risk since course corrections cannot be made safely without visual cues. Infrared imagery can improve visibility in degraded visual environments, however when heavily degraded, this method alone is insufficient. Obscurants and noise can be rapidly categorized and suppressed in real-time by having an artificial intelligence trained to select and apply the most suitable modern image processing technique for a given frame. Phase I efforts have completed a proof-of-concept and successful prototype. Phase II efforts will deploy this to a testbed system for testing. Additional data from new environments will be acquired and analyzed. New image restoration methods will be added and the AI revised to utilize them. Dual-use algorithms will be explored to enhance driver performance. Hardware acceleration and embedded system design will also be studied to improve the testbed further. Phase III efforts will focus on widespread implementation.