Study On Tobacco Maturity Rapid Detection Method

Tobacco maturity is affecting the quality of tobacco leaves, and the key factors affecting the commercial value of tobacco.This paper will use the rapid detection methods based on near infrared spectrum, dielectric properties, chlorophyll content and color features, to study the rapid detection methods of tobacco maturity, provide theoretical basis and technical support for the rapid identification of tobacco maturity and quality control;And develop a rapid detection of tobacco maturity android software based on color features, installed on the small mobile devices such as mobile phone, pad and so on, can be real-time, fast, accurate judgment of tobacco maturity, rapid and accurate judgment, has the vital significance to the timely collection and scientific modulation.The main research contents and conclusions are as follows:(1)Qinyan96 tobacco leaf as the research object,to study the relationship between the spectral characteristics of tobacco leaf and tobacco maturity.and set up BP neural network tobacco maturity discriminant model based on near infrared spectrum,the average recognition rate of lower, middle and upper tobacco maturity reached 97%, 91% and 81%.(2)Qinyan96 tobacco leaf as the research object,to study the relationship between the dielectric properties(such as equivalent capacitance Cs,loss coefficient D and the equivalent impedance |Z|)of tobacco leaf and tobacco maturity.and set up BP neural network tobacco maturity discriminant model based on corresponding dielectric parameters,the average recognition rate of lower, middle and upper tobacco maturity based on equivalent capacitance Cs reached 85%, 48% and 41%,the average recognition rate of tobacco maturity based on loss coefficient D and the equivalent impedance |Z| is lower than equivalent capacitance Cs.(3)Qinyan96,Yunyan99,K326 tobacco leaf as the research object,to study the relationship between the chlorophyll content of tobacco leaf and tobacco maturity.and set up BP neural network tobacco maturity discriminant model based on chlorophyll content,the average recognition rate of lower, middle and upper K326 tobacco maturity reached 75%, 92% and 75%,but the average recognition rate of Qinyan96 and Yunyan99 is slightly lower than K326.(4)Yunyan99,K326 tobacco leaf as the research object,to study the relationship between the color features of tobacco leaf and tobacco maturity.To explore the accurate characterization of ripening color patterns and characteristic parameters of tobacco maturity; on this basis, On this basis, to develop a rapid detection of tobacco maturity android software, the recognition errors of Yunyan99 and K326 tobacco maturity are less than 0.5.