Visible Electro-Optical (EO) System and LIDAR Fusion for Low Cost Perception by Autonomous Ground Vehicles

The Toyon Research Corporation and Missouri University of Science and Technology (Toyon-MST) team proposes development of a system which fuses multi-spectral EO and 3D data collected with a low-cost LIDAR to enable improved perception and navigation for unmanned ground vehicles (UGVs). Phase I research and development will include system design based on analysis of sensor and algorithm capabilities using measured EO and LIDAR data, as well as feasibility demonstration of critical algorithm components. The proposed development includes innovative algorithms for fully-automated dense 3D reconstruction from 2D EO images, with accurate representation of intermediate reconstruction ambiguities. This enables near-optimal calibration of the EO and LIDAR sensors, as well as LIDAR sensor tasking and data fusion to obtain high-resolution 3D information using a low-scan-rate/low-cost LIDAR. The proposed development also includes innovative algorithms for fused EO-LIDAR obstacle detection, and leverages the experience of the Toyon-MST team with UGV platform, sensor, and algorithm development. The proposed Phase I work is designed to set the stage for development and demonstration of a complete prototype and real-time processing implementation in Phase II, and for integration of the developed technology in Navy systems.