Classification of Medicinal Plant Using an Optimized Deep Learning Method

Convolutional Neural Network or CNN is a popular approach in deep learning commonly used in Image Recognition because of its high and optimal potential in deep learning. In this study, CNN is used to classify common backyard medicinal plants that can be found in the Philippines which can be further integrated into mobile applications for use by people in remote locations where medicinal plants serves as first aid remedies. There are training images belonging to three categories of medicinal plants which are distributed in training, validation, and testing and are fed to a CNN with 3 Convolution Layer with Rectified Linear Function (ReLU) Activation, 3 Max Pooling Layer, Flatten, and Dense Layer with Softmax Activation, and Adam as the set optimizer. The CNN provides outstanding performance in training so as to validation which makes it a good model to implement in such a context.