SUPPORT VECTOR MACHINE BASED CLASSIFICATION FOR TOMATO LEAVES DISEASES

Abstract

Tomato is a very common vegetable used by Indians. It is widely harvested by farmers all over India including
Odisha. But often, the yield quality and quantity are affected due to various diseases. In this study, the focus is
to design an automated leaf disease detection technique based on image processing and machine learning for the
early diagnosis of infection in tomato leaf. The model is designed based on a dataset containing 18,200 images
of tomato leaves those classified in 10 classes (9 diseases and a healthy class) collected from the Plant Village.
For feature extraction, methods like shape-based features, color-based features, and texture-based features are
used, then different machine learning algorithms like linear regression, Decision tree, Random forest, and
Support Vector Machine (SVM) are used for tanning and testing the model. Among all SVM is fitting very well
with the underlying dataset and producing a classification accuracy of 98.2%