THE COMPARISON OF MATCH FILTERING AND SPECTRAL INFORMATION DIVERGENCE METHODS FOR DETECTING OF WATER QUALITY

Abstract

The quality of water resources, which are very important for environmental and human health, can now be examined together with remote sensing technology. Management of water resources is crucial for the development of countries. In our country, the quality of the water resources is getting worse due to the development of the industry and other factors. There are many spectral classification methods in remote sensing, and the applicability of two spectral classification methods is aimed to evaluate water quality in this study. Within the scope of this study, Imranli region of Kizilirmak River was examined. High resolution satellite image in operation CHRIS Proba is provided free of charge from European Space Agency. The spectroradiometer measurement points were determined to obtain the water sample. From these points, local measurements were made with the spectroradiometer simultaneously with transition of the satellite. The quality evaluation was carried out by analyzing the water samples in the laboratory. As a result of measurements made with the spectroradiometer, reflectance from the measurements was resampled at the Chris PROBA band intervals and a reference for classification was established. Using this data, applicability of a spectral classification method, Match Filtering and Spectral Information Divergence, was tested in order to evaluate water quality.