Automatic Detection Of Wheat Flour Precision Based On Image Processing

Wheat flour contains unique ingredients, so wheat flour was given a wide range of uses. In our daily diets, pasta occupies a large proportion, and wheat flour can produce a wide variety of food, other food crops are incomparable. Wheat flour is one of the most important components of the modern agricultural product processing industry. In the process of wheat flour, processing accuracy of wheat flour is higher, the bran content is lower, and the color is whiter, so the wheat flour has higher the commodity value. Therefore, the yield and quality of wheat flour are of affected by the processing accuracy of wheat flour. At the same time the market price of wheat flour and the economic benefits of wheat flour processing enterprises are also directly affected by the of processing accuracy. At present, China's current national standards "GB5504-85 wheat flour processing precision inspection method" is a kind of sensory inspection method, wheat flour was processed by hand for ride powder or steamed bun, the visual result was compared with the pink and bran of physical standard samples to determine the accuracy of wheat flour processing. This approach not only influenced by subjective factors, but also complicated operations, the judgment errors are larger, the result statements are inaccurate, and is not conducive to connect with international standards. Along with technological development, people have developed such as white-meter, colorimeter, color grade device, near-infrared instruments to measure wheat flour processing precision. However, these instruments are research one of the characters of wheat flour, wheat flour can not be comprehensive assessed from many areas of the characteristics of wheat flour processing precision.The subject is to solve the questions of the national standard wheat flour testing methods according to the actual situation of enterprises, and look for the best indicators based on digital image processing technology, and research and development the new detect method base on digital image processing, Wheat flour classifiers have been successfully developed. This method establishes a kind of simple, rapid, accurate, objective of wheat flour processing precision detection methods for the majority of the wheat powder enterprises, and to some extent, to shorten the detection time, saving the cost of the purpose of testing. Course of the study is divided into the following parts:First of all, extraction the image's color characteristics of wheat flour. Comparison of various common feature extraction method, the final option of using the CIE L * a * b * color model to extract the color features of wheat flour. Because it is a kind of simplified uniform color system, high brightness, very suitable for smaller color measurement and situations. Through verification, CIE L * a *b * color space can effectively extract the 12 features of wheat color.Next, the OTSU algorithm extracted the gluten content of stars the image in various grades of wheat flour. The 120 Wheat flour images contents of bran were extracted .Finally, the use of choice principle in recent samples of fuzzy pattern recognition identifies unknown type of wheat flour image.