Enhanced Plant Leaf Disease Detection Using Deep Learning Techniques

Enhanced Plant Leaf Disease Detection Using Deep Learning Techniques

DR.R.Gopi1, R.Aarthy2, Kothamasu Venkata Ratna Sai3, Jasthi Manikanta4, Chunduri Sai Babu5, Bommisetti Pardha Sai6

1 Associative Professor, Department of Computer Science and Engineering, Dhanalakshmi Srinivasan Engineering College, Perambalur

2 Assistant Professor, Department of Computer Science and Engineering, Dhanalakshmi Srinivasan Engineering College, Perambalur

3,4,5,6 Student, Department of Computer Science and Engineering, Dhanalakshmi Srinivasan Engineering College, Perambalur

Dhanalakshmi Srinivasan Engineering College (AUTONOMOUS),Perambalur-621212

Abstract: Indian economy is largely depending on the growth in the agricultural production. Nowadays, iden-tifying the plant leaf disease is one of the most challenging tasks in the agricultural sector because plant leaf diseases spread very quickly and effects the yield of the crop. In this paper, we proposed three deep learning models i.e, VGG16, VGG19 and InceptionV3 for detecting various plant leaf diseases because manual detection sometimes leads to inaccurate results. The dataset consists of several variety of plants like tomato, apple, potato, sugarcane, grapes and corn of both healthy and diseased leaves and all the images are collected manually and also from various freely available resources. Using the three models we achieved an average accuracy as 91.8% accuracy which demonstrates the feasibility of the profound deep neural network approach

Keywords:
VGG16, VGG19, InceptionV3, deep  learning, Convolution Neural Network (CNN).