Fruit Classification System Using Multiclass Support Vector Machine Classifier
Mohammed Zeeshan, Akshatha Prabhu, Chandni Arun, N. Shobha Rani
Abstract
Fruit classification has always been a challenging task as it involves sorting, grading, and analyzing fruit quality specifically for import, consumption purposes and for workers at the Point of Sale (POS) in supermarkets. The proposed work has used computer vision and support vector machine (SVM), for classification. Among the fruit images collected, 505 fruit images are used in training and 150 fruit images are used in testing. The work is carried out in various stages. First, the input image is resized to 256×256 resolutions. Later, pre-processing is performed using Gaussian filter to enhance the image quality by reducing the noise. Feature space is created by extracting the colour, texture and shape features. Then, principal component analysis (PCA) is applied to reduce the dimensions of the feature space to overcome the curse of dimensionality. Further, the support vector machine (SVM) classifier is used for training the data. Total of 655 images distributed across 18 categories of fruits are maintained that achieves 87.06% accuracy.