Pengenalan Jenis Bunga Menggunakan Principal Component Analysis Dan Jaringan Syaraf Tiruan

Abstract

Flower has variety of species and shapes. In the flower species recognition, classification is a difficult task because of close shape similarity among different flower classes. Any flowers that have similar shape are usually grouped into the same flower class. However, different species of flowers can have shape that look similar to one another. Lighting conditions and viewpoints when the flower image was taken, also can be varied. All of these circumstances obviously lead to a confusion among the flower classes when flower images of the different flower species that have look similar shape were classified. This research applies the Principal Component Analysis and Backpropagation Neural Network Method in flower species recognition system. Objective of this research is to find out the best flower species recognition that can be used to identify flower image. This research involves 480 flower image data and 8 species of flowers, which consists of 280 training images and 200 testing images. From the results of this research, the Principal Component Analysis and Artificial Neural Network Method shows the good performance of flower species recognition system, with an accuracy average is 97%.
Keywords: flower species recognition, Principal Component Analysis, Artificial Neural Network, Backpropagation