Fruit Disease Detection and Segregation using Machine Learning

• Published in: 2024 5th International Conference on Image Processing and Capsule Networks (ICIPCN) pp. 304 - 309
• Main Authors: Shanthini, E., Swetha, N., Taaha Arshath, M.R., Thrinethra, M.S.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 03.07.2024
• Subjects: Computer vision; Convolutional Neural Network; Disease Detection; Image processing; Machine learning; Machine learning algorithms; Manufacturing industries; OpenCV Image Processing; Packaging; Pre-processing; Production; Segregation
• DOI: 10.1109/ICIPCN63822.2024.00057

Agriculture serves as the backbone of the country, and all agricultural products, particularly perishable items, are mass-produced, packaged, and marketed by industries. Among all commodities, fruit and vegetable packaging requires the utmost attention, with only the unaffected ones packed. It is not assured that all fruits and vegetables are packed in their healthiest state. While packaging, even when one of the fruits or vegetables is contaminated it likely spreads to the others. This may widely affect the production while exporting. This huge-scale operation requires quick action to prevent the spreading of diseases and therefore detect the diseased fruits and vegetables using a Machine Learning (ML) algorithm- Convolutional Neural Network (CNN) by comparing the trained data with the test data, thereby providing a more precise means than traditional approaches. However, this way of employing a machine learning algorithm, CNN has been limited to the detection segment alone. Eventually in the manufacturing industries, segregation of fruits and vegetables was not implemented automatically. Using the existing machine learning method for detecting diseases in fruits and vegetables as a reference, this study has proposed a new method for automatic segregation of unaffected fruits from the diseased ones by using a hybrid Computer Vision (CV) and CNN based image processing technique.