<![CDATA[Classification and prediction of tomato maturity level are automated using the Backpropagation Neural Network method. The determination of maturity in agriculture is still applied manually. With the development of technology in the field of image mining, determining the maturity of tomatoes can be done automatically. The method used in making this system is an Artificial Neural Network. The algorithm used is backpropagation. The output of tomato ripeness consists of three categories, namely immature, half ripe and ripe. 60 training data and testing data were used. The backpropagation architecture in this study consists of 3 input layers, 4 hidden layers, and 1 output layer. The activation function used from input to hidden layer is binary sigmoid, while from hidden layer to output is the identity function (purelin). Image extraction in the form of RGB minimum value is useful as input. Processed and produces output maturity level and maturity prediction. The results of testing the training data system obtained an accuracy value of 96.67% and testing data of 90%.]]>
Copyrights © 2020
