Published in Scientific Papers. Series E. Land Reclamation, Earth Observation & Surveying, Environmental Engineering, Vol. IV
Written by Levent BASAYIGIT, Rabia ERSAN
Determination and classification of plant patterns with satellite images is most suitable method for preparation ofsource inventory. The use of high resolution satellite imagery is quite widespread for separation of land use types. Pixelbased classification methods are widely used to classify images. The choice of classification methods in satellite imagesdirectly affects the success of the research. In recent years, the use of object-based classification method to determinethe plant pattern as parcels is being investigated. In this study, we present an example to determine the crop pattern asparcels using pixel and object based classification methods.The study was conducted in Isparta-Turkey. Quickbird-2 satellite image, ERDAS and e-Cognition Developer Trial 8.8software were used. Maximum likelihood and Isodata algorithms of Pixel-based classification methods and,multiresolution segmentation operator of object-based classification methods were used to classify satellite imagery.Object-based classification method was determined to be more successful than pixel-based classification method, basedour analyses.
[Read full article] [Citation]
image classification
