Research About Quality Grading Of Wheat Grain Based On Improved AHP

Wheat grain quality is an important indicator of test wheat varieties, the grain grading technology can be precise quantitative indicates the degree of wheat quality, it not only has important reference value for wheat breeding work, but also on the morphological characteristics of wheat grain quality to identify them. Using modern computer image processing technology for quality grading of wheat kernels with nondestructive, classifying the advantages of high accuracy, high speed, it can replace artificial labor huge, repeatability. Aiming at wheat grain quality grading, this paper in the grain as the research object, through the research method based on digital image processing has carried on the research on wheat grain quality grading method.Traditional on wheat grain quality classification methods are usually based on artificial observation is given priority to, workers rely on experience to evaluate wheat grain, with significant limitations and uncertainties, this paper, by using digital image processing technology using mathematical modeling methods, through the wheat grain image acquisition, preprocessing, to extract the wheat grain in the perimeter, area, plumpness, rectangle degrees, color saturation, stretched length6eigenvalues, statistical analysis, the feature database is established. Comprehensive use of analytic hierarchy process (AHP) and the subordinate function method realized on wheat grain quality classification, and the realization of the algorithm using Matlab and Vc++, to verify the effectiveness of the algorithm. In this paper, the main work is as follows:1. The grain image acquisition technology researchIn order to guarantee the grain image acquisition effect, different weather, illumination and time was studied, and finally to select natural weather good period, with the grain color contrast the big black board as the background color, and adopts the method of vertical shooting, collected the grain image.2. The grain image preprocessingIn wheat grain image preprocessing, according to the grain image gray distribution, through the different threshold of image enhancement. In the image smooth, median filter algorithm and neighborhood were used respectively to average algorithm effect comparison, finally used the median filtering algorithm with better effect denoising. In image segmentation, edge function in Matlab to contour extraction of wheat grain. Finally through image grayscale, the image enhancement, image smoothing, image segmentation, image binarization, image pretreatment process to get the grain synthetically.3. The eigenvalue extraction and selection of wheat kernelsThe processed wheat grain image, extraction of the shape, size, color, texture, etc multiple characteristics, through the calculation, finally selected the perimeter, area, plumpness, rectangle degrees, color saturation, stretched length6characteristics as recognition characteristics.4. Improve the grain classification model based on analytic hierarchy process (ahp) On wheat grain quality grading method in the study, first of all, according to the grain characteristics of six parameter information is simple and effective analytic hierarchy process (ahp) is adopted to improve the quality grading, the grading after the misjudgment of the wheat grain quadratic discriminant according to membership function and comprehensive evaluation model is concluded. Through the experiment, obtained the satisfactory results, to prove the effectiveness of the method.