APLIKASI PROGRAM MATLAB UNTUK ANALISA CITRA HYPERSPECTRAL PADA AKAR DARI TANAMAN KELAPA SAWIT YANG MENGALAMI KEKURANGAN AIR

Abstract

Hyperspectral images are three dimensional images which have two dimension spatial information and one  dimension spectral information. Hyperspectral image processing using Matlab program is preferable because it is more adaptive for many analysis purposes. This research was aimed  to construct Matlab to process and analyze the hyperspectral images of the roots of oil palm plants that have experienced water deficiency. The program was designed and constructed using a GUI . The use of a GUI aims to combine each pixel of the same line from each sample to produce a new image. The samples were roots  of oil palm plants that experienced simulated water deficiency by giving different water volumes of 0 mL, 1000 mL, 2000 mL and 3000 mL (normal). The optical method used in this study is a hyperspectral imaging method which has 650 nm diode laser  as the light source , spectrograph Specim Imspector V10 , and a  monochrome CMOS as a detector. Reflectance intensity versus wavelength  was extracted from each images and analyzed. The results showed that the Matlab GUI program that had been constructed was able to produce 1024 new images that had a pixel size of 15× 1280 from each sample. The results also show that the reflectance intensity values are higher at higher water deficiency of the oil palm roots.