Klasifikasi Biji Jagung Berdasarkan Tekstur Dan Warna Menggunakan Metode Backpropagation Berbasis Web

Citra

Abstract

Indonesia is an agrarian country because most people rely on the agricultural sector for their livelihood. Paiton sub-district of Probolinggo district in East Java is one of the majority of people who are farmers until now still cultivating corn crops. Corn is widely consumed by the surrounding community because it is rich in nutrients, corn can also be used as food and animal livestock. Therefore, the quality of corn quality must be maintained in such a way because corn production is decreasing in productivity every year due to the reduction of planting land. Gray Level Co-occurrence Matric (GLCM), RGB (Red, Green, Blue) and Backpropagation methods. So the researcher will use this method to classify the quality of corn kernels. It is hoped that this utilization can solve the problem of middlemen so as not to lose money when buying corn from farmers. The result of this research is the process of determining the quality of corn kernels based on color and texture features using the GLCM, RGB, and Backpropagation methods with a total data of 150 images consisting of 120 training data and 30 testing data. The results of testing the classification system obtained an accuracy value of 75%. So that the backpropagation method can determine the quality of corn kernels based on images using a computer system so that it can be implemented into Web Flask.