부산대학교 도서관

There are currently around 375000 known species of plants in the world. Expert botanists are able to easily classify and classify them based on either division/phylum, order, class, species, genus or family. But common people like students and new people in the field may find it difficult to classify them appropriately due to lack of experience or exposure to those plants. In the proposed solution, we plan on suggesting a system which would use deep learning models for image processing. This system can be trained on an ample amount of plant leaves images and tree images dataset containing pictures of various plant leaves and trees. Many prominent datasets like Flavia, Swedish Leaf datasets, etc can be used to train the model. The model will be built using a combination of 2-D Convolutional layers, Max Pooling and Dense layers. The system will take the images of the plant or tree as input and the built model will work on and predict the output. The output will be the classification of the plant.