Comparison between HSV and YCbCr Color Model Color-Texture based Classification of the Food Grains

This paper presents the comparative study among HSV and YCbCr color models in the classification of food grains by combining color and texture features without performing preprocessing. Also, the paper deals with the effect of training set, block size and K-value in the process of classification. The proposed method is performed in two phases; the feature extraction phase and classification phase. The K-NN and minimum distance classifiers are used to classify the different types of food grains using color, local and global features. The classification of food grains involves the computation of features locally and globally using the Haralick features and the cumulative histogram respectively. The non-uniformity of RGB color space is eliminated by HSV and YCbCr color space. The good classification accuracy is achieved using both the color models. General Terms Pattern Recognition, Algorithm, Features, Classification.