DEEP LEARNING BASED AUTOMATED DETECTION AND CLASSIFICATION OF TOMATO LEAF DISEASES

Abstract

Tomato is one of the world's most consumed crops and highly profitable. However, its production process is affected by several leaf diseases leading to low quality and yield. It is important to detect these diseases at an earlier stage to minimize losses and boost productivity. The study focuses on building an automatic plant disease detection system based on deep learning. CNN algorithms are used in analyzing images of tomatoes' leaves and classifying them to specific types of plant diseases like bacterial spots, early and late blights, leaf mold, and also healthy leaves. Annotated images are fed into the machine to enable it to classify the input into one of these types with accuracy. Resizing, normalizing, and augmentation procedures are carried out on image data for increased effectiveness and precision. The resulting model classifies leaf images with very high accuracy and minimal human effort. Experiment results prove that the suggested model outperforms other traditional machine learning models in terms of efficiency and accuracy of prediction. Such a model can be useful for farmers in disease detection while minimizing dependence on manual diagnosis and improving farming practices. In general, the study advances technology in smart agriculture.