<![CDATA[At this time, the leaf authentication method is widely used in the classification process of herbal plants. Basically, the leaf authentication method compares the image to be identified and the reference image created in the dataset. This paper aims to identify the leaves of herbal plants using an artificial intelligence method, namely Convolutional Neural Network (CNN) that is embedded on Raspberry Pi. CNN has an advantage that it does not require feature extraction, because in CNN, automatic feature extraction already exists. This paper uses seven types of leaves from different herbal plants. Leaf images are taken using a camera and processed by Raspberry Pi, which is integrated with CNN. Identification was carried out on seven types of herbal plants divided into two-thirds of training data and one-third of testing data. The identification process results will be validated with other data not included in the training data and testing data, as well as leaf data other than the seven types of leaves identified. The CNN method shows good results in the authentication process, with an accuracy rate of 93.62% for testing data offline and 91.04% for testing data online.]]>
Copyrights © 2020
