Medicinal Plant Processing Viewing Leaf Morphological Feature Extraction With Artificial Neural Tissue

Abstract

Indonesia has long known and used plants that are efficacious as medicine. Of the many medicinal plants in the world, 80% of medicinal plants grow in tropical forests in Indonesia. About 28,000 plant species grow and 1,000 species of them have been used as medicinal plants. The large number of medicinal plant species and the high degree of similarity can cause errors in the process of identifying medicinal plant species. So it takes the help of a computer to identify the types of medicinal plants. The purpose of this study was to identify types of medicinal plants using backpropagation neural networks based on leaf morphological feature extraction. The results show that changes in the learning rate value affect the results of the identification of medicinal plant species based on leaf morphological features. The results of the calculation of the average recognition rate value of 90% for training data and 75.56% for testing data occur when the learning rate is 0.01. The best learning rate value for the identification of medicinal plant species is 0.01 with an average number of epochs of 11.67 and MSE of 0.13. This shows that leaf morphological feature extraction methods and backpropagation neural network algorithms are very well used to identify medicinal plant species.