Detection of Multiple Plant Diseases by Leaf Diagnosis Using Deep Learning

India is a land of agriculture, and it has a deep impact in the economic growth of our country. Around 10-20% crops in India, suffer from various diseases that can lead to those crops getting wasted therefore this is one of the major concerns for the farmers to protect their crops from such diseases. To prevent the plants from getting wasted the disease should be detected as soon as possible so that the farmer can take the necessary steps to save the crops from getting wasted. Farmers have been using their conventional methods for years to identify diseases, but it is hard for them to identify various diseases with high accuracy. Now with the advancement in Artificial Intelligence and Computer vision, we can use various techniques to solve this problem of disease detection in different plants with more accurate results. This will lead to a fast and early diagnosis of the disease and will help to decrease crop wastage making it more beneficial for the farmer and for the whole country. In this research paper, we have made use of Deep Learning a very popular technology nowadays as it provides various models and functionalities to work with images. We have used the Convolutional Neural Network (CNN) model for this research paper as it is one of the best in image classification and is also good at extracting smaller details from the image which will work well for the detection of disease in a plant by diagnosing the images of its leaf. We have made use of Plant Village Dataset for our model which contains more than 60,000 RGB images of both healthy and unhealthy leaves as well.