Grading of quality in tomatoes using multi-class SVM

Abstract

Tomato is one of the major crops cultivated across the globe. Despite the hype, the crop is also suffered from wide variety of diseases. Early detection of these disease will proportionally increase the crop yield and consequently improves the economic growth and food reserves of a nation. In particular the leaf part is more prone to diseases. Based on the leaves and by using mutli-class SVM we can even classify whether the plant is in healthy or unhealthy condition. An accuracy of 99.1 % is achieved on three classes of diseases-early and late blight and mosaic.