Plant classification using deep learning

Abstract

For protecting biodiversity, species knowledge is very essential. The classification of the species by the general information
is complex, time consuming and frustrates for the non-experts. Although plant classification based on the multi organ is
also very arduous, where the added information provides can help to discriminative the species by the means of shape and
appearance of the plant will also raise the ramification of the problem. There are many existing techniques which rely on
hand-crafted features and also increasing the process of automating the species classification. In this paper, a deep learning
method is employed to extract discriminative features to be learned for the plant images; here the convolutional neural
network (CNN) involves both the generic and organ information combining using the new feature scheme for the
classification of the plant species. Furthermore, we compare methods based on classification accuracy exploit on
plantclef2015benchmark.In computer vision our results are admissible to researches in ecology for their ongoing research
trail.