Plant species recognition methods using leaf image: Overview

Plant plays an important role in agricultural, industrial, medicine, environmental and ecological protection. Recently, with global warming, biodiversity loss, rapid urban development and environmental damage, people have been seriously destroying the natural environments, which results in that a large number of plant species constantly dying and even dying out every year. It is essential to protect plant species. The first step of protecting plants is to recognize them and understand what they are and where they come from. But there are a large number of plant species that have been named on Earth, and many are still unknown yet, it is difficult to identifiy each species. To handle such huge information, develop a quick and efficient classification method has become a significant research. Plant species can be recognized by its leaf, flower, skin, fruit and seed, etc. Relatively speaking, using leaf to recognize plant species is very simple and convenient, and many leaf based plant species recognition methods have been proposed. In this paper, we mainly summarize the existing leaf based plant species identification methods, including plant leaf characteristic, public databases, feature extraction based methods, subspace learning based methods, sparse representation based methods, and deep learning based methods. The aim is to emphasize the importance of plant species identification, train people to know about plant species, and provide guidance and comprehensive study for the beginners in this field, in turn, to treasure and protect plant species.