Classification of flour types based on PSO-BP neural network

In order to improve the accuracy of recognition of flour types based on image processing, this paper presents a method of identifying flour types by PSO-BP neural network based on image processing. In this paper, firstly, the average values of red, green, blue and gray arc extracted by color histogram. Then, the normalized RGB (three -primary colors) space model is converted into the HSV (color) space model, and the luminance components are extracted from the HSV model. Secondly, comparing the two - and three - dimensional characteristic classification map, the average value of the grayscale of the flour image, the blue average value m the three stimulus values and the ratio of the brightness and the blue mean value are selected as the feature quantity. Finally, the different kinds of flour are classified by PSO-BP neural network. The experimental results show that the method can be used to identify the varieties of flour quickly and accurately, which provides a new idea for the detection and identification of flour types.