APPLICATION OF IMAGE ANALYSIS TECHNIQUES IN SPACE TECHNOLOGY FOR DETECTING THE ENVIRONMENTAL DISASTER

Abstract

In this work image processing techniques were developed for identifying the burned and unburned surfaces. Both thresholding techniques and support vector machines were explored for application with satellite data from NOAAIAVHRR ((National Oceanic and Atmospheric Administration! Advanced Very High Resolution Radiometer) and MODIS (Moderate Resolution Imaging Spectroradiometer). The thresholding algorithm was designed to remove the water bodies from the land coverage. This algorithm also highlights the hotspots or fire spots in the suspected regions. While support vector machine was used for training to remove the false points which were also highlighted in threshold technique. Both the approaches can be used separately or in combination depending on the size of the image. The RBF (Radial Basis Function) kernel was used in training of three set of inputs: surface temperature, NDVI and GEMI respectively. This makes a classified image in the output highlighting the burned and unburned surfaces. The results obtained were compared with the fire hot spots, and found to be reasonably corresponding to the result obtained. This project has been implemented using C++ in Linux along with use of GRASS and GLADE packages. Satellite image processing packages like ENVI and ERDAS were also used at various stages of the project