A Distributed Algorithm for Critical Area Detection in Satellite Imagery

This paper presents a distributed algorithm for critical area detection in satellite imagery. The detection of critical area is very important in natural disaster detection and control. The algorithm uses the distributed change detection methods for multiple spectral bands in satellite images and combines changes which have appeared during a period of time. The result shows a critical area for analyzed time interval. The algorithm is distributed for support a large amount of data. The distributed approach shows a decrease in processing time with the increase of the number of workers. The presented algorithm represents a component from MedioGRID system, which implemented a real-time satellite image processing system for extracting relevant environmental and meteorological parameters on Grid systems. The critical area detection application will be integrated in Grid environments.