SCAFLE - Self Calibrating, Adaptive, Fused LIDAR/EO Sensor

The ability to fuse electro-optical (EO) imagery with range data enhances the robustness of unmanned ground vehicle (UGV) autonomy by allowing discrimination between types of terrain that cannot be distinguished on the basis of geometry alone (non-traversable mud vs. dry soil, non-traversable rocks vs. traversable brush, etc.). Performing this fusion requires knowing both the intrinsic (optical center, focal length, etc.) and extrinsic (relative offset and orientation) calibration parameters of the sensors. These can vary over time as sensors are replaced due to damage, the system is jolted during use, etc. In order to be practical for warfighter use, such a fused sensor system needs to be able to self-calibrate using features of the environment, and to adaptively update that calibration as shock or other sources of variation cause it to change. Such adaptive self-calibration avoids the need to train warfighters in calibration procedures, and eliminates the need for additional equipment. We propose designing a compact, low-cost Self-Calibrating (Adaptive) Fusion of LADAR and EO imagery (SCAFLE) sensor that will provide fused visible multi-spectral EO and LADAR range data covering a full 360 degree field of regard to support perception capabilities such as terrain classification and visual odometry.