Semi-Automated DIRSIG scene modeling from 3D lidar and passive imagery

The Digital Imaging and Remote Sensing Image Generation (DIRSIG) model is an established, first-principles based scene simulation tool that produces synthetic multispectral and hyperspectral images from the visible to long wave infrared (0.4 to 20 microns). Over the last few years, significant enhancements such as spectral polarimetric and active Light Detection and Ranging (lidar) models have also been incorporated into the software, providing an extremely powerful tool for multi-sensor algorithm testing and sensor evaluation. However, the extensive time required to create large-scale scenes has limited DIRSIG’s ability to generate scenes ”on demand.” To date, scene generation has been a laborious, time-intensive process, as the terrain model, CAD objects and background maps have to be created and attributed manually. To shorten the time required for this process, this research developed an approach to reduce the man-in-the-loop requirements for several aspects of synthetic scene construction. Through a fusion of 3D lidar data with passive imagery, we were able to semi-automate several of the required tasks in the DIRSIG scene creation process. Additionally, many of the remaining tasks realized a shortened implementation time through this application of multi-modal imagery. Lidar data is exploited to identify ground and object features as well as to define initial tree location and building parameter estimates. These estimates are then refined by analyzing high-resolution frame array imagery using the concepts of projective geometry in lieu of the more common Euclidean approach found in most traditional photogrammetric references. Spectral imagery is also used to assign material characteristics to the modeled geometric objects. This is achieved through a modified atmospheric compensation applied to raw hyperspectral imagery. These techniques have been successfully applied to imagery collected over the RIT campus and the greater Rochester area. The data used include multiple-return point information provided by an Optech lidar linescanning sensor, multispectral frame array imagery from the Wildfire Airborne Sensor Program (WASP) and WASP-lite sensors, and hyperspectral data from the Modular Imaging Spectrometer Instrument (MISI) and the COMPact Airborne Spectral Sensor (COMPASS). Information from these image sources was fused and processed using the semi-automated approach to provide the DIRSIG input files used to define a synthetic scene. When compared to the standard manual process for creating these files, we achieved approximately a tenfold increase in speed, as well as a significant increase in geometric accuracy