Identification method of particle material heterochromatic object based on color feature

Color feature and its measurement are important foundations and means in evaluating product quality and are widely used in the food, pharmaceutical, industry and agriculture fields. The paper implements the identification of heterochrome objects in particle materials based on color feature. According to color model classification from the view of application, it selects RGB color model and HSI color model to extract color features. Then it uses Ohta color model, considering the influence on identification effect caused by the correlation of each color model component. The paper selects histogram, color matrix and wavelet packet transform to extract color features. It compares the features extracted by every color model and identifies them using threshold method. Experimental results indicate that color features extracted by HSI color model have the best effect on identifying heterochrome object.