Comparison of vegetation indices in distinguishing crops with similar spectral characteristics: Sanliurfa, Ceylanpinar

Each object interacts differently with energies of different wavelengths. These energies are absorbed, transmitted or reflected in different amounts by objects. All these properties are called spectral properties of objects. Thanks to these features, objects on the earth can be distinguished from each other by remote sensing. In the study, a solution to the problem that crops with similar spectral characteristics cannot be distinguished with high accuracy using single-time images was investigated. In this context, a region of 30 parcels under the use of Ceylanpinar Agricultural Enterprise, located in Ceylanpinar district of Sanliurfa province, was chosen as the study area. GNDVI (Green Normalized Difference Index), TGI (Triangular Greenness Index), NDRE (Normalized Difference Red Edge Index), CIRedEdge (Chlorophyll Index) and MCARI (Modified Chlorophyll Absorption in Reflectance Index) index images were created with a Sentinel-2 satellite image detected when wheat and lentil crops grown in this region were in their high chlorophyll-containing period. Then, these images were classified and the differences between each other were compared. In the accuracy analysis, 210 randomly distributed control points were used for all images. According to the results, all classifications performed gave an overall accuracy of over 80%, and the highest overall accuracy was obtained in the CIRedEdge image with 98.57%. Evaluations made; shows that indices containing red edge bands in the formula are a factor that makes a difference in distinguishing crop species when they are in their high chlorophyll-containing stages.