Identifying the Quality System of Cocoa Beans to Increase Productivity Using Backpropagation Neural Network Algorithm: A Case Study at Sumber Rejeki Farmers Group, Patuk Gunung Kidul

Abstract

An image is a two-dimensional function (x, y) consisting of a number of finite elements. Image processing uses Backpropagation and Principal Component Analysis (PCA) Artificial Neural Networks. The processed data is the image data of 256 x 256 pixel cocoa beans taken from the Sumber Rejeki Farmer Group (KTSR) with the results of two types of identification output, namely quality cocoa beans and non-quality cocoa beans. The identification process in the system through two stages of the process, namely the training process where the image data of cocoa beans through the training process with input parameters ANN. Image data that has been completed through the training process is used for the testing process. Based on the results of the study using the Backpropagation method with parameters alpha value = 0.35, tolerance value error = 0.05, iteration value = 1000, input neuron value = 255, hidden neuron value = 6, output neuron value = 3 produce quality image data 14 and 6 quality images that are not qualified from the test image data are 20 images of cocoa beans with an accuracy rate of 70% and an error rate of 30%. Based on the tests that have been conducted, it can be concluded that the ANN Backpropagation algorithm can produce a classification of cacao seed quality with a high degree of accuracy.