Author(s)

SAIF UDDIN, A. AL-DOUSARI & A. AL-GHADBAN

Abstract

Quickbird panfused data with 60 cm resolution is used to map the locations of date palm trees in the arid land of Kuwait. In this study, Laplacian maxima filtering was applied to classify date palm trees using high-resolution satellite imagery. The processing was done in two steps: the first step involved smoothing of the data using non-linear diffusion and the second was extracting local spatial maxima of Laplacian blob used for palm tree identification. The results are promising and the classification accuracy in the two test areas is 96% and 98%, which is higher than maximum likelihood classification for the same dataset. The results show that this methodology can be adopted for the mapping of palm trees in arid Middle Eastern countries.

Keywords

arid land, blob, image classification, laplacian filtering, quickbird