Abstrait

Comparison of remote sensing techniques to exploit irrigated agricultural lands from satellite images

Alikhah-Asl Marzieh, Forootan Elham, Namdar Mohammad

Knowledge of agricultural and horticultural land use percent is necessarily important for supplying human food and should be considered in agricultural planning. Remote sensing provides valuable data on land use classes. Mapping land use through remotely sensed images comprises various considerations, processes and techniques. In this research, four methods (ICA, Tasseled Cap, NDVI and supervised classification) have been utilized for extraction of irrigated land class in a part of Hablehrood watershed. The results of this investigation showed that among the studied techniques; ICA which uses the higher order statistical characteristics of multispectral and hyper spectral imagery such as skewness and kurtosis has the highest accuracy, whereas supervised classification has the lowest accuracy. Moreover; this research revealed NDVI accuracy is more than Tasseled Cap.