The Digital Imaging and Remote Sensing Image Generation (DIRSIG) model has been developed at the RochesterInstitute of Technology (RIT) for over a decade. The model is an established, first-principles based scene simulationtool that has been focused on passive multi- and hyper-spectral sensing from the visible to long wave infrared (0.4 to 14 µm). Leveraging photon mapping techniques utilized by the computer graphics community, a first-principles based elastic Light Detection and Ranging (LIDAR) model was incorporated into the passive radiometry framework so that the model calculates arbitrary, time-gated radiances reaching the sensor for both the atmospheric and topographicreturns. The active LIDAR module handles a wide variety of complicated scene geometries, a diverse set of surface and participating media optical characteristics, multiple bounce and multiple scattering effects, and a flexible suite of sensormodels. This paper will present the numerical approaches employed to predict sensor reaching radiances andcomparisons with analytically predicted results. Representative data sets generated by the DIRSIG model for a topographical LIDAR will be shown. Additionally, the results from phenomenological case studies including standard terrain topography, forest canopy penetration, and camouflaged hard targets will be presented.
Department, Program, or Center
Chester F. Carlson Center for Imaging Science (COS)
Laser Radar Technology and Applications X 5791 (2005) 342-353
RIT – Main Campus