Cinnamomum Camphora Classification Based on Leaf Image Using Transfer Learning

Plants play a very important role in our daily life providing us with food and oxygen. It is necessary to build an aut omatic system for recognizing plant Convolutional neural netw orks (CNNs) have become effective identification method especial ly in computer vision in recent years because they achieved outst anding performance on different tasks, such as image classificatio ns. In this paper, we have shown performance comparison of usin g the seven pretrained deep convolutional neural network(Alexne t, Googlenet, Resnet50, Resnet101, Inceptionv3, Inceptionresnetv 2, Squeezenet) that employed fine-tuning technique to retain the l ast three full connection layers and get the classication model. We applied our method to Cinnamomum Camphora leaf dataset clas sification,.the experimental results showed that the proposed met hod is quite effective and feasible.