Classification Quality of Tobacco Leaves as Cigarette Raw Material Based on Artificial Neural Networks

Determination of the current tobacco grade classification performed by the tobacco commonly called grader with a variety of human frailties. Therefore it is necessary to develop classification automation tools. But earlier experiments need to be done first, in this case using Backpropagation Neural Network classification approach.From this research was obtained increased accuracy for the classification grade tobacco leaf with Backpropagation Neural Network method obtained an accuracy of 77.50%. This indicates that the feature extraction parameters such as shape, color, and texture applied to a Neural Network Backpropagation method can produce a level of accuracy that is quite accurate. Tests were also carried out to produce a level of precision and recall satisfactory as well. Based on the data testing eksperimet of 40 tested for classification grade tobacco leaf there are 8 different datasets that result accuracy between Backpropagation Neural Network with a grader.