Nutrition Diagnosis Of Wheat Based On Image Processing

In order to solve the complex diagnosis of traditional nitrogen nutrition,environmental pollution caused by excessive fertilization,and the problem of land zoning,this paper used image processing technology to carry out research on the nitrogen nutrition diagnosis model of wheat.The experiment used Jimai No.22 in daejeon as the research object,using the Sony FDR-AXP35 4K HD camcorder to capture the image of the wheat canopy leaves.Some preliminary research results were obtained through the study of the correlation between the color characteristic parameters based on the SPAD value(chlorophyll value)and the value of the N value,the prediction and estimation of the color characteristic parameters and N values of the population of wheat.The specific work is as follows:(1)Research on image data acquisition and image preprocessing.Using the SONY FDR-AXP35 4K HD camera to obtain wheat leaf images at the jointing stage of wheat,a white background plate was used to block it.From the obtained images,the image quality is high and the background is simple,which is conducive to image processing and feature extraction in later stage.In addition,this paper also uses the threshold segmentation method to divide the wheat leaf color,and preserves the color information of the wheat leaves,which lays the foundation for the extraction of the color feature and the feature analysis in the later period.(2)The relationship model of color characteristic parameter and N value based on chlorophyll value is established.Firstly,through the linear correlation analysis,the color characteristic parameters which are related to the SPAD value are selected as B,R,G,G and R,and the regression model between the color characteristics and the chlorophyll value SPAD is established by using the linear regression model.Then,the correlation between chlorophyll content gradient and nitrogen value was further analyzed,and the relationship between SPAD value and nitrogen content of wheat was estimated.With the help of chlorophyll SPAD,a mapping relationship between color characteristics and nitrogen content was established.Finally,through the model test,the root mean square error of the model is 0.16,the relative error is between 11%~27%and the accuracy is 0.69,which shows that the estimation achieves high accuracy.(3)The prediction model of color characteristic parameter and N value of group wheat was established.Firstly,based on the image information,four color feature parameters are selected:standard NRI for red,NGI for green,NBI for blue,ratio R/G for red and green.Then,linear correlation analysis and linear regression analysis were used to establish 4 relational models of color characteristic parameters and N prime values.Finally,an independent test sample isused to assess the model’s estimation ability,and the relative error E,the root mean square error RMSE and the coefficient R2(coefficient of determination)are used to evaluate the estimation ability of the model.The estimated root mean square error of the model is less than0.050,the relative error is between 0.54% and 7.35%,and the accuracy is 0.875,indicating that the estimation achieves higher accuracy.