Remote Sensing Image Processing Based on Modified Fuzzy Algorithm

Abstract

The study focused on the problem of natural objects image segmentation using intelligent technology. As the natural objects for this research the authors chose tundra vegetation. Concerned organizations often monitor hard-to-reach remote natural and technological objects using aerospace surveillance tools. Multi- and hyperspectral data provide them with automated image segmentation and assessment of the state of landscape elements. At the same time, the processing of space imagery is a difficult task which often goes under uncertainty. The formation of a training sample for image processing algorithms is a component of the integrated processing technology of heterogeneous data. The aim of the work is to present a monitoring technology for natural objects that applies the mathematical apparatus of fuzzy logic to automated processing of multi- and hyperspectral aerospace imagery. The technology is likely to be used in regional and industry-specific decision support systems for managing complex natural and technological objects.