Plant Pathologist- A Machine Learning Diagnostician for the Plant Disease

Crop infection in terms of bacteria, fungus and other viruses is an unrelenting issue and loos of food due to this is a key challenge faced by farmers for many decades. Proper identification of this disease may prevent this loss. This study describes precision agriculture in terms of a mobile application called ‘plant pathologist’ for detecting plant diseases that forecast their occurrence using machine learning and artificial intelligence. A combination of computer vision and deep learning algorithms were used to evaluate images of plant leaves and compared them to a database of known diseases in order to make a quick and accurate diagnosis. The development process involved the identification of the research issue and the objectives, the gathering and pre-processing of data, the choice and training of the model, and the model evaluation through this quantitative methodology. The convolutional neural network (CNN) was trained on a sizable dataset of annotated plant disease photos as part of the machine learning model utilized in the mobile app. The CNN was trained to recognise the visual characteristics of photos of healthy and diseased plants and classify them according to several disease types. To ensure adaptability and precise identification of plant diseases, the app has been trained on a specific type. With hyperparameter tuning and cross-validation, the best-performing CNN model was chosen for the model selection procedure. The mobile application, which offers thorough information on plant disease diagnosis is then updated with the chosen model. This offers farmers, botanists, and plant enthusiasts an invaluable resource to accurately identify plant diseases through machine learning.