Quality Analysis of Rice Grains Using Morphological Techniques

Abstract

Human beings from all over the world prefer and consume rice more than any other food. Rice is at its peak of demand when its quality is good. Currently, the kind and quality of rice is determined through a naked-eye visual assessment approach. This method, however, is arduous, time-consuming, requires human skill, and is dependent on the inspector's physical health. To address these issues, this work introduces an automated system that uses digital image processing techniques to identify and classify rice grains. The image processing approach is the most appropriate since it is a non-contact technique that captures the picture of the rice grains. MATLAB is used to pre-process, segment, and extract features from the captured images. Using Neural Networks (NN) and Support Vector Machines (SVM) algorithmic classifications, the quality of rice is assessed based on the extracted features. The results indicate that SVM-based classification performs better than its counterpart based on our comparison study.