Mung Bean Variety Classification Using KNN and Image Processing Technique
Jesusimo L. Dioses
Abstract
Mungbean is a rapidly maturing crop that plays a vital role in crop diversification. It serves as a cash crop, intercrop, and rotation crop with rice. Additionally, it is one of the most affordable protein sources in the Filipino diet, rich in potassium, B vitamins, magnesium, and fiber. San Mateo, Isabela, known as the "Munggo Capital of the Philippines," has long utilized the "munggo after rice" farming strategy. Identifying mung beans manually can be tedious and requires expertise, with accuracy potentially compromised by factors such as poor vision or color blindness. This study developed a model to identify three different mung bean varieties: the Chinese Variant, Kulabo, and Pagasa 7. Using canny edge detection in image processing and a supervised machine learning approach with a KNN classifier, the model achieved an accuracy of 93.1% with 5 neighbors.