Leaf Classification Based on Convolutional Neural Network. Issue 1 (March 2021)

• Record Type: Journal Article
• Authors: Wu, Peng; Qian, Zhou
• Is Part Of: Journal of physics. Volume 1820:Issue 1(2021)
• Publication Date: 2021-03
• Subjects: Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/1820/1/012161
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: Convolutional Neural Network (CNN), a very important neural network structure in deep learning, is a network model often used in image classification, target recognition and other fields. In botany, leaf classification and recognition is very important for identifying new or scarce tree species. In nature, plants are widely distributed, and the survival and development of all living things on the earth depend on plants. Identification of species by leaves and related research are of great help to the study the evolution law of plants, the protection of plant species and the development of agriculture. This paper uses convolutional neural network in artificial intelligence to identify the leaves of several kinds of trees collected by Kunming Institute of Botany, Yunnan Province, which can realize the automatic extraction of leaf image features, reduce tedious labor costs, and realize the use of artificial intelligence to classify leaves, thus providing an auxiliary means of artificial intelligence for botany research.