Vol. 12, Issue 12 (2023)

Author(s):
Anurag Shrivastava

Abstract:
Potato is one of the most widely consumed crop in India, also there is significant very high demand of potatoes globally. However, some potato diseases are the leading cause of declination of potato quality and production. Diseases, such as early blight, late blight, common scab, potato leafroll virus etc. have a significant impact on the production quality and quantity, an automated system based on computer vision can help the farmers for early detection of leaf diseases. This study proposes a machine learning based approach for early detection of potato leaf diseases. Overall 2150 samples have been observed from online database. The overall samples are divided into three class: early blight, late blight and healthy and divided into three groups (Group1: early blight and healthy, Group2: late blight and healthy and Group 3: early blight and late blight). Various machine learning algorithms has been employed for classification of early blight, late blight and healthy potato leafs. An highest accuracy of 98.4% has been achieved using convolution neural network (CNN) for late blight nad healthy group, indicating the feasibility of the machine learning techniques and may help to reduce losses incurred by the potato farmers due to crop diseases.