Multiple Fruits Classification Using Computer Vision

Abstract

Fruits are an abundant source of vitamins, minerals, and fiber. There are numerous varieties of fruit, including apples, lemons, guavas, gooseberries, oranges, berries, melons, tomatoes, and avocados. Fruit classification would be put to use in industrial application. In this paper, Our proposed method consists of four main steps: object detection, image processing, feature extraction and Classification. Collected a dataset of multiple fruits images captured in a black background. Six categories of fruits are included in the dataset such as apple, lemon, orange, guava, gooseberry, and tomato. To train and test the random forest classifier, divide the dataset into training and testing sets. The accuracy of the proposed approach for classifying multiple fruits using computer vision techniques was 75 percent on the training set and 81 percent on the testing set, according to the results, which show its effectivity.