Multicriteria selection of a method for processing multispectral Earth remote sensing data

Abstract

The article is devoted to the use of qualimetry methods for models and poly model complexes in order to solve one of the relevant engineering problems-automation of selecting methods for calculating Earth remote sensing (ERS) data processing when analyzing the state of complex natural and technical systems. The proposed approach was discussed using the example of choosing methods for calculating forest sustainability indicators. A typical situation was considered when alternative methods and models can be applied at each stage of data processing. The essence of the proposed approach is to formulate and solve the task on multi criteria comparative analysis of processing methods based on a set of indicators, which include costs, required for implementation one or another method, efficiency, which refers to calculation duration of the analyzed sustainability indicator, and an indicator reflecting the quality of the solution-accuracy of calculations result. The solution algorithm was illustrated within an example of choosing the method for assessing consequences of the forest fire. The selection results were presented in the form of a table, which allows the user to evaluate losses and gains in the values of partial indicators when moving from applying one method to another. The proposed algorithmization of the selection task determines possibility for its automation and, thereby, simplifying application of complex methods for processing ERS data for the end user. In addition, the possibilities and degree of validity for scaling the results of processing ERS data from individual areas to large forest areas are expanding.