Cognitive Processing and Exploitation of 3D Laser Imaging Detection and Ranging (LIDAR) Imagery Data

Toyon Research Corporation proposes the research and development of advanced cognitive algorithms for search, localization, and target classification in wide-area 3D LiDAR point-cloud data. The proposed algorithms seek to extend and scale recent advances made in 1D and 2D cognitive signal processing to support the automated analysis of 3D LiDAR PCD using deep learning and neural-filtering architectures. The proposed algorithms are based on a model of human cognition, which involves coarsely processing data to identify regions of interest (ROI), and subsequently increasing the granularity of detection to resolve target signatures of interest. The proposed effort includes the development of a high-performance, low-latency cognitive cortex, including supporting analysis and visualization routines that demonstrate a complete end-to-end prototype of the developed cognitive wide-area search, localization, and automatic target recognition algorithms.