Optimal Bands selection for Soil Classification and Moisture Mapping
Study of feature selection algorithms with application to soil classification and estimation of soil moisture
Versandkostenfrei!
Versandfertig in 6-10 Tagen
32,99 €
inkl. MwSt.
PAYBACK Punkte
16 °P sammeln!
Study of soils and their moisture content is one such area where Remote Sensing has wide applications. However, study of soils was not attractive until the advent of Hyper-spectral/Multi- spectral Imaging by satellite sensors, which provides reflectance information across the different contiguous spectral bands. The large amount of information in these bands served as features in the classification of soils. A spectral library of the different type of soils at different moisture levels serves as a reference for the purpose. However, a satellite sensor is designed to meet multiple criteria, whi...
Study of soils and their moisture content is one such area where Remote Sensing has wide applications. However, study of soils was not attractive until the advent of Hyper-spectral/Multi- spectral Imaging by satellite sensors, which provides reflectance information across the different contiguous spectral bands. The large amount of information in these bands served as features in the classification of soils. A spectral library of the different type of soils at different moisture levels serves as a reference for the purpose. However, a satellite sensor is designed to meet multiple criteria, which restricts the number of spectral bands that can be used and their resolution. Moreover, not all features contribute to classification. Hence, it becomes mandatory the optimal combination of the spectral bands is selected for a satisfactory classification of soils and the estimation of their moisture content. This work compares the performance of a hybrid algorithm, made by combining two sub-optimal methods, with the performances of the other classical algorithms for feature selection.
Andere Kunden interessierten sich für
