3-DIMENSIONAL PRE- AND POST-FIRE COMPARISON OF FOREST AREAS
Keywords: Forest fire, VHR satellite imagery, orientation, image matching, DSM generation, 3D change detection
Abstract. Forest agencies give special attention to forest fires where post-disaster loss can rarely be gauged in a quick and economic way unless an appropriate technology is adopted. Determination of the planimetric and volumetric changes between pre- and post-fire is in high demand. The FORSAT system (a satellite processing platform for high resolution forest assessment) was developed to meet the relevant demands. It has the capability of extracting of 3D geometric information from the very-high resolution (VHR) imagery from satellite optical sensors and automatic 3D change detection. FORSAT includes two main units. The first one is dedicated to the geometric and radiometric processing of satellite optical imagery and 2D/3D information extraction. This includes: image radiometric pre-processing, image and ground point measurement, improvement of geometric sensor orientation, quasi-epipolar image generation for stereo measurements, digital surface model (DSM) extraction by using a precise and robust image matching approach specially designed for VHR satellite imagery, generation of orthoimages, and 3D measurements in single images using mono-plotting and in stereo images. FORSAT supports most of the VHR optically imagery commonly used for civil applications: IKONOS, OrbView – 3, SPOT – 5 HRS, SPOT – 5 HRG, QuickBird, GeoEye-1, WorldView-1/2, Pléiades 1A/1B, SPOT 6/7. The second unit of FORSAT is dedicated to 3D surface comparison for change detection. It allows users to import DSMs, to co-register them using an advanced 3D surface matching approach and to calculate the planimetric and volumetric changes between epochs. The capacity and benefits of FORSAT have been tested in two real cases, where are burned areas located in Cyprus and Austria. The geometric characteristics of burned forest areas have been identified both in 2D plane and 3D volume dimensions, using pre- and post-fire optical data from different sensors. FORSAT is a single source and flexible forest information solution, allowing expert and non-expert remote sensing users to monitor forests in three and four dimensions from VHR optical imagery for many forest information needs.