Object Based Classification of Crop Pattern Using Multi-Temporal Satellite Dataset in Multi-Cropped Agricultural Areas: Lower Seyhan Plane Case Study

Abstract

Lower Seyhan Plane (LSP) is one of the most productive agricultural basins of Turkey and it covers main part of the Çukurova Region. The crop productivity in the study area is much more than most developed countries and Turkey’s average productions. Ideal spatial conditions such as climate, soil and transportation for agriculture creates these productive lands. The aim of this research was to define winter and summer crop pattern using multi-temporal Landsat satellite dataset applying object based classification technique. Crop pattern was detected according to 2013 hydrological term (October 2012 – September 2013) as winter and summer. Landsat dataset was defined according to the greenest and cloud free times of the crops. Object based classification was applied because of regular parcel distribution of the crops. As a result of the study; general kappa coefficiency of LSP was obtained as 0.9. According to the results, it was found that while wheat, potato and onion for winter crops were determined as areal distribution, corn and cotton as first crop and corn as second crop in summer season. © 2017, Centenary University. All rights reserved.