Pendugaan Reflectance Berdasarkan Citra Digital Menggunakan Principal Component Analysis untuk Menduga Usia Daun Jati Belanda

Abstract

Jati Belanda (Guazuma ulmifolia) is one of medicinal plants that contained various chemical components, one of them is flavonoid. Because of Jati belanda has many benefits in health sector, this plant was studied to control the quality of flavonoid in different ways, usually, chemical lab and tools are used for that experiment. Such process has high accuracy and fast, but it can damage the sample and required very high cost. Therefore, this paper discussed on development of a system that has function as quality control method to control the quality of medicinal plant leaf with spectral image using principal component analysis (PCA) and probabilistic neural network (PNN) is used to determining age of leaf based on value of estimated reflectance. The result of vale of estimated reflectance using PCA has RMSE value of 15.26 and GFC of 91.42%. In this research, identification system is builded by using polynomila order 2 and the first four components to predicting age of Jati Belanda leaf. Identification system has an accuracy of 86.67%.