Modern approaches to processing large hyperspectral and multispectral aerospace data flows

We consider approaches to processing large hyperspectral and multispectral imaging flows produced in aerospace monitoring for solving a wide range of problems of management of natural resources, environmental security, prevention of natural disasters and technogenic accidents, as well as problems of real economy, and basic and applied sciences. We analyze the specific features of the phases of hyperspectral data analysis and describe a software and hardware system that uses new and improved methods and algorithms for processing large flows of hyperspectral and other aerospace data and has a high-performance computer. This system contains different types of software for identifying the types of given objects by solving inverse problems of remote sensing as well as by analyzing their qualitative and quantitative characteristics, combined multiparameter processing of hyperspectral aerospace data, tracking the local changes including those related to changes in meteorological conditions and vegetation periods, detecting and identifying the types of small objects on the basis of analysis of individual parts of the image, detecting and identifying heat sources, etc. We bring examples of processing of hyperspectral and multispectral satellite images with the help of software and hardware tools developed.