Hybrid algorithm in identifying the plant disease and to suggest medical advice. Issue 4 (July 2021)

• Record Type: Journal Article
• Authors: Roopa, D; Murukesh, C; Rayavel, P; Renukadevi, B; Eswaran, Malathi
• Is Part Of: Journal of physics. Volume 1964:Issue 4(2021)
• Publication Date: 2021-07
• Subjects: Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/1964/4/042026
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: Agriculture sector is the most important which is the reason for the life of every human being and every living objects in this world. The top tier problem of all formers in agriculture is to recognize the disease affected in a particular crop and their causes. Some disease spread easily and quick that may lead to disaster in the targeted production of that particular period. To overcome this problem, it would be best to recognize the diseases affected at initial stage. There are many systems exist for recognize the disease but these systems are concentrate only on the particular parts of the plant such as leaf, steam etc, and more over they did research using artificial neural networks (ANN) and some of the paper using support vector machines algorithm (SVMs). By using these algorithm, they can only detect whether it is affected or not and also they can concentrate only on the particular parts of the plant such as leafs. But this is not efficient. To overcome this demerit in our proposal we are aimed to concentrate on the entire plant for this purpose we are using the combination of following algorithms recurrent neural networks (RNNs), support vector machines (SVMs) and random forest algorithm. By using these algorithm, we can design a system which is more efficiently identify the plant's diseases compare to the previous systems which is based on the image(plant) provided by the end user after identification process we provide complete information about the plant's current condition and what type of pesticide should be used and how long it to used. In our system we are also plan to provide periodic intimation to the end user for the image of the affected plant for the purpose of tracking the diseases condition and provide complete agri assistant until the plant completely recovered from the disease. Our proposal mainly concentrates on the formers who are all not having sufficient knowledge about the plant disease, this system will save money and time of the formers.