Detection Of Tobacco Baking Quality And Design Of Control System Based On Machine Vision

Tobacco industry is one of the most important cash crops in China,plays an indispensable role in the national economy.However,the international competitiveness of China’s tobacco products is insufficient.At present,the tobacco baking room system has unsatisfactory quality of baking,low automation level,consumes a lot of manpower and material resources,and has high demands on the professional requirements of flue-cured tobacco workers.Therefore,the urgent need to find a baking method to solve the question,improve the quality of tobacco baking,the liberation of labor.In this study,Shaanxi Hanzhong Yunyan 99 tobacco was studied,based on the appearance characteristics of tobacco leaves during curing process.The machine vision technology is used to collect the tobacco image information and extract the image characteristic parameters,to explore the correlation between appearance characteristics and internal quality of tobacco leaves during baking,and to establish a prediction model based on appearance characteristics and internal quality,An automatic control device based on the change of appearance characteristics of tobacco in baking process was designed.(1)The research of tobacco image information acquisition based on machine vision.Build the tobacco image acquisition system,and determine the model of each part of the tobacco image acquisition system,it mainly includes cameras,computers,light sources and camera black box.Get the tobacco image during baking,and preprocess it,gray level processing,FFT transform,mean filter and edge detection.The color characteristic parameter values of tobacco leaves were extracted in three color spaces of RGB,Lab and HSI respectively,and the change curves of color parameter values with baking time were drawn,find a fire point,draw a baking process curve,as the theoretical basis of the control system.(2)The relationship between internal quality of tobacco leaves and their appearance characteristics during baking.The results showed that the change of color parameters and the degradation tendency of pigment were basically the same.The main changes were in the "yellowing" period,"color fixing period" and "dry gluten stage",which changed slowly.The water content of tobacco leaves decreased with the increase of baking time,and the range of decrease was different in different periods of baking.The morphological change of tobacco leaves showed an increasing trend with the increase of baking time.The yellowing stage increased rapidly,the color setting stage increased slowly,and the length,shrinkage and width shrinkage of the dry gluten stage were basically unchanged.The correlation analysis shows that during the curing process,the correlation of tobacco leaf water content,chlorophyll and carotenoid between pigment and the color parameters,leaf length and width shrinkage shrinkage of the most significant or extremely significant.Through the regression analysis of the internal quality of tobacco leaves and the external characteristics of tobacco leaves,a linear regression model of chemical components and color parameters can be established,Through the color characteristics of tobacco leaves,the content of chemical components in leaves can be rapidly predicted,and the technological curve of baking can be further optimized,It provides a theoretical basis for the design of grilled room control system based on image information.(3)A new type of oven control device based on image information is designed and developed,The hardware design of the device mainly includes the main control module,information acquisition and processing module,drive circuit and so on.The system software design including the main program design,program design of serial communication,the image and the temperature and humidity information acquisition and processing program,LCD display program design,program design and key alarm circuit.Finally,we test the hardware circuit of the dense baking room control device based on the image information,through constant modification.The device can basically realize the automatic control of the baking room based on the image information,but the overall accuracy of the control device needs to be improved,and the function is still to be improved.