Comparison of Accuracy Rate for Soybean Type Using Backpropagation

Abstract

Abstract – Soybean is a pods plants that were use a base component to make food such as tempe and tofu. Soybean has a few kind of shapes, circle, circle and flat, oval, oval and flat. Soybean also has a few kinds which is white soybean, black soybean and green soybean (edamame). On this article the accuration level from recognation of those 3 kinds will be compared based on the portrait distance. There were three distance, 10 cm, 15 cm and 20 cm. Feature extraction that used is GLCM and recognation using backpropagation algorithm. The result after preprocessing, feature extraction with GLCM and recognation with backpropagation algorithm with the portrait distance of 10 cm accuration level is 100%, 15 cm accuration level is 99.07% and 20 cm accuration level is 77.22%Keywords: Soybean, GLCM, Backpropagation