NDVI Optimization Using Genetic Algorithm

Applying ratioing on multispectral images is one of theimportant techniques used in remote sensing. In this paper aformula containing multiplication, addition and a division at theend is proposed to calculate a measure for classifying land covers.Applying this formula on each pixel of the multispectral imagetogether with a set of thresholds, one can decide if class of thepixel is vegetation, water, soil, etc. NDVI is used to determinewater, vegetation and soil areas in a map. In this article UrbanLandsat images are used as the experimental dataset. GAoptimization is used to derive the optimized coefficients for akind of NDVI formula. Finally, results are evaluated to see if theoptimized formula is more accurate and more robust than thetraditional NDVI.