Research And Design Of Tobacco Leaf Grading System Based On ARM

Tobacco leaf is a special agricultural economic crop in my country.From planting to harvesting,it is carried out under strict national management.The flue-cured tobacco that is finally baked by farmers will be acquired by the state.As the world’s largest tobacco consumer,China has a huge annual demand for tobacco leaves.The state has a classification system for acquisition of flue-cured tobacco,price of flue-cured tobacco varies greatly depending on the grade.However,the existing classification method still relies on human senses and experience,which is very slow and subjective.Therefore,a fast,accurate and convenient classification method has a wide range of application value.In order to achieve rapid and accurate classification of tobacco leaves,this paper designs a tobacco leaf classification system based on the current national standards and digitizes tobacco leaf information.The specific content is as follows:1.Design and build the hardware system of ARM as the main control module,complete the hardware selection and debugging of the main control module,display module,image acquisition module,weighing module and other modules,and run normally.2.According to the existing national standards as the theoretical basis,the characteristic factors required for the classification of tobacco leaves in this paper are proposed,and the value of the characteristic factors and the threshold value of each characteristic factor are determined.Among them,the method of predicting the edge function of tobacco leaves based on the fitting calculation method,and the method of calculating the area of the tobacco leaf by the integral calculation method,compared with the traditional method of calculating the damaged area in the image,reduces the impact of the damage of the edge of the tobacco leaf on judging the damage of the tobacco leaf.influences.The support vector machine algorithm with voting mechanism is used to determine the color classification.Compared with the single use of the support vector machine algorithm,the accuracy rate is increased from 89.8% to97.2%.3.The model of tobacco leaf grading was established.The method of combining pruning strategy with decision tree algorithm was used to create the hierarchical model,and the accuracy of the hierarchical model was increased from 89.1% to 91.6% compared with the hierarchical model created by decision tree algorithm alone.4.Design the software of the tobacco leaf grading system,use Python language as the writing language to complete the software of the grading system,and the designed related functions can run correctly after being transplanted to the hardware system.