Spectral Reflectance Estimation Based On Digital Leaf Image Using Wiener Estimation for Sambiloto Leaf Age Prediction

Abstract

This research proposes a new method to estimate the spectral reflectance for Sambiloto (Andrograp/1is paniculata) leaf age prediction based on leaf digital image. Sambiloto is a medicinal plant containing andrographolide compounds. Wiener estimation is used to estimate spectral reflectance based on RGB values and probabilistic neural networks (PNN) is used to classify plant age. Analyses of the result showed that the number of terms in the Wiener estimation affects on the results. According to experimental result, ten terms gave the best result for spectral reflectance estimation and accuracy of leaf age prediction is 73.61 %. This prediction can be used as quality marker of medicinal plants.