IDENTIFICATION OF RICE VARIETIES USING MACHINE LEARNING ALGORITHMS

Rice, which has the highest production and consumption rates worldwide, is among main nutrients in terms of being economical nutritious our country as well. Rice goes through some stages from field to dinner tables. The cleaning phase separation rice unwanted materials. During classification phase, solid ones broken are separated calibration operations performed. Finally, process extraction based on color features, striped stained other than whiteness surface grain separated. In this paper, five different varieties belonging same trademark were selected carry out using morphological, shape features. A total 75,000 images, including 15,000 for each varieties, obtained. images pre-processed MATLAB software prepared feature extraction. Using a combination 12 4 features 90 obtained spaces, 106 extracted images. For classification, models created with algorithms machine learning techniques k-nearest neighbor, decision tree, logistic regression, multilayer perceptron, random forest support vector machines. With these models, performance measurement values sets 12, 16, 106. Among success average accuracy was achieved 97.99% morphological 98.04% It regression 99.25% 99.91% perceptron When results examined, it observed that addition new feature, increases. Based obtained, possible say study classifying varieties.