DECOMPOSITION OF SPECTRAL FEATURES OF REMOTE SENSING BASED ON THE COMPONENTS OF THE CORRELATION COEFFICIENT
A. V. Lapko1,2, V. A. Lapko1,2, S. T. Im3,2
1Institute of Computational Modelling, Siberian Branch, Russian Academy of Sciences, Krasnoyarsk, Russia
2Reshetnev Siberian State University of Science and Technology, Krasnoyarsk, Russia
3Sukachev Institute of Forest, Siberian Branch, Russian Academy of Sciences, Krasnoyarsk, Russia
Keywords: decomposition of statistical data, automatic classification, correlation coefficient, remote sensing data, spectral data analysis
Abstract
A method is proposed for decomposing the range of values of two-dimensional spectral features according to the values of their constituent correlation coefficients. The basis of the technique is the analysis of the product of the normalized values of spectral features. The peculiarity of the indicator used and the thresholds entered by the user for its values make it possible to decompose the initial statistical data and map the results obtained. Unlike traditional methods, the proposed approach has higher computational efficiency, which is necessary when processing large amounts of statistical data. The results of the application of the technique in processing of remote sensing data of a natural object are considered.
|
