Open AccessIdentification of Sugarcane Foliar Diseases Methods and Datasets
Author(s) -
Swapnil Dadabhau Daphal,
S. M. Koli
Publication year - 2020
Publication title -
international journal of engineering and advanced technology
Language(s) - English
Resource type - Journals
ISSN - 2249-8958
DOI - 10.35940/ijeat.c6454.029320
Subject(s) - cash crop , identification (biology) , chlorosis , agriculture , convolutional neural network , crop , rust (programming language) , sugar industry , plant disease , computer science , agricultural engineering , microbiology and biotechnology , artificial intelligence , sugar , agronomy , geography , biology , engineering , botany , programming language , biochemistry , archaeology
Agriculture is the major part of the Indian economy as it provides key support to social and economic development of the country. Sugarcane is the leading cash crop in various states of India which has larger share in the net agriculture produce. Recently, researches have highlighted the impact of different disease on various plants. Estimated loss is much severe for the sugarcane crop via foliar diseases. Foliar diseases like rust, eye spot, mosaic and banded chlorosis may hamper the overall productivity of sugarcane and sugar recovery rate (RR). Early prediction of these diseases may limit the losses in terms of produce net benefits. This paper addresses the concerns, types of the foliar disease and researches undertaken to overcome the problems related to the diseases. Morphological characters of these diseases may help in identifying the representative features and to use them for the optimum classification. Currently the use of deep neural networks (DNN) is encouraged for the classification. DNN demands the huge and accurate databases. Intuitively the use and important methods used in database creation for disease diagnostic system (DDS) has been highlighted in the paper. Modifications made to the Convolutional Neural Network architecture have suggested the improved performance in terms of recognition accuracy (RA) and lesser recognition time.