An Automatic Plant Disease Detection System Using Deep Learning Technique

Authors

  • A. Mallikarjuna Reddy, Salma Samreen Shaik

DOI:

https://doi.org/10.17762/msea.v71i3.121

Abstract

Plant diseases generate significant losses in terms of productivity, economics, quality, and quantity of agricultural produce, among other things. The fact that the majority of the Indian population is reliant on agricultural output indicates that it is necessary to manage the loss caused by plant diseases. In order to prevent such illnesses, it is necessary to monitor the plants from the beginning of their life cycle. The usual form of supervision for this is by naked eye observation, which is more time-consuming and requires a high level of skill. As a result, in order to accelerate this procedure, we need an autonomous plant disease detection system that can both identify and deliver cures for infections. Beginning with the collection of photographs in order to establish a database, the necessary steps for putting this concept into action are laid out below. It is possible to do image preprocessing, which is used to load and train the datasets of an image, as well as classification by using Multi-class classification in Support Vector Machine (SVM), since this technique trains the samples of two or more classes in the same amount of time.KNN algorithm at the training phase just stores the data ,it will predict features and identify whether it is healthy or unhealthy .We build a model for CNN for detecting the disease through the datasets and compared the algorithms with CNN model to find out the algorithm performance of each algorithm. By training CNN model we obtain the train and validation accuracy and loss curves. The Experimental model show the accuracy of SVM with 94 %,KNN with 82% & CNN with 96 %.

Downloads

Published

2022-06-08

How to Cite

Salma Samreen Shaik, A. M. R. . (2022). An Automatic Plant Disease Detection System Using Deep Learning Technique. Mathematical Statistician and Engineering Applications, 71(3), 152–158. https://doi.org/10.17762/msea.v71i3.121

Issue

Section

Articles