QualiOn: An Onion Quality Analyzer and Sorter Device using Random Forest Classifier

TLDR: In this paper , the authors used a random forest algorithm to identify onion's external disease and its cause and sort the healthy onions from the diseased onions using 200 images of red and white onion.

Abstract: Onion is one of the most important crops grown in the world and is generally used as vegetables, spices, and medicines. The inspection of onion quality relies on the visual part of the crop, most known diseases are based upon two things: the leaves and the roots. When either of the two is damaged, it is automatically a bad one. When harvest time comes, manual sorting is already implemented at the field. The study aims to identify the onion's external disease and its cause and sort the healthy onions from the diseased onions. Dataset preparation was a collection of 200 images of red and white onions. The images were augmented to increase the diversity of data for machine learning training and testing. The trained model using random forest algorithm were implemented in raspberry pi 4 which act as the main controller of the system. The system was tested for 30 trials and yield an accuracy of 96.67% in identifying the onion disease and sorting based on its quality.