Partice Swarm Optimization based fusion of MODIS and PALSAR images for hotspot detection

SAR images have the capability to detect and identify various features of lands on the basis of scattering behavior of the targets. It is observed that various features like soil moisture, roughness, crop health and crop growth etc., can be easily detected with SAR, but detection of subsurface fire (hotspot) with SAR images has to be explored. Therefore, in this paper we have explored the possibility of detection of hotspots with fully polarimetric SAR images. Several polarimetric indices like CPR (Cross Polarization Ratio), HV/HH and HV/VV have been extensively studied on PALSAR (Phased Array type L-band Synthetic Aperture Radar) data to check the sensitivity of detection of subsurface fire by which proper indices can be selected to detect hotspots. Many researchers were using high resolution optical and thermal images for hotspot detection. But, still several challenges exist. It is known that SAR and Optical images are providing some complementary information. Therefore, in this paper, Particle Swarm optimization (PSO) based image fusion technique has been proposed to detect hotspots. SAR and Optical data i.e. MODIS (Moderate Resolution Imaging Spectroradiometer) data are used for fusion purpose. Before fusing the MODIS, several indices like MSAVI (Modified Soil Adjusted Vegetation Index), PAVI (Purified Adjusted Vegetation Index) etc. indices were tested for their sensitivity to detect hotspots. A quite good result has been observed after fusion of SAR and MODIS data.