Material Classification for Physics-Based Sensor Simulation Using Stereo-Pair Imagery

In order to generate geospecific materialization and textures for synthetic scenes derived from satellite based stereo imagery, a complete model or description of the reflectance is required for surfaces in that scene. Surface reflectance accounts for various characteristics the human eye uses to distinguish objects in natural or image scenes. Attributes include: material type (metal, plastic, vegetation etc.), shape, specular versus diffuse reflection, glossiness and surface texture are examples. Reflectance is typically characterized by the Bidirectional Reflectance Distribution Function (BRDF) which describes the differential change in spectral radiance reflected (dLr (?r??r)) off a surface with respect to incident spectral irradiance dE(?i??i) to that surface. SimWright Inc. applies twelve years experience in remote sensing, 3D visual simulation, stereo photogrammetric image processing (including tool development) and systems integration to propose a remote sensing approach to measure BRDF for physics-based material classification and synthetic texture generation. The approach is built upon extensive airborne and ground based testing (model validation) which will measure target objects’ spectral radiance in elevation and azimuth. Skydome irradiance will also be characterized. Accurate BRDF models can be incorporated into physics based modelers to generate 3D terrain surfaces and objects for IGs.