| Matcha is a micropowder tea product that processed by steaming,drying and then grinding by the mortar or modern ball mill,which usually uses the fresh shoots of the tea plantation as raw material.Currently,the detection of the quality grade of matcha in most processing enterprises still rely on the traditional physics and chemistry analysis and artificial sensory evaluation methods to test the internal and external qualities of matcha.However,this method usually is time-consuming and costly,which has some certain constraints for the quality of the matcha rapid detection.In this case,the research of rapid nondestructive detection method and portable equipment development based on the visible/near infrared spectroscopy technology has been carried out in this paper.The main contents and conclusions are as follows:1.Study on rapid detection of the color of matcha using the visible spectrocopy technology.In this study,a rapid method based on visible spectroscopy technology has been attempted to detect the color of matcha.The quality grade of matcha was evaluated according to the value of L*a*b*.Firstly,the raw spectral data of the matcha in the range of 380~780nm were received the useful integral wavelength,and then the XYZ tristimulus values were counted using the equivalent wavelength instead of the integration and were converted to the required values.Secondly,the accuracy of the method was verified by 45 samples of different grades of matcha,using this method and standard chromometer,the correlation of the L*a*b* were observed to higher than 0.9.The results showed that the method can meet the requirements for rapid detection in actual production.Then,using this method to detect the L*a*b* values of different grades of matcha samples and studying the change regulation of different grades of matcha in the L*a*b* color space.On the basis of this,the model of grade evaluation based on the color of matcha was established successfully by using the L*a*b* values of 60 grade qualified matcha samples combined with euclidean distance method.Finally,the accuracy of the model was evaluated with 60 different grades of matcha samples,the accuracy rate was 86.67% compared with the result of artificial evaluation experts.The overall results showed that the method based on the color can be used as the reference for rapid detection of matcha grades.2.The design of rapid non-destructive testing equipment for detecting color of matcha.According to the characteristics of matcha processing technology,a rapid non-destructive detection device that mainly includes hardware systems and corresponding software systems was developed to quickly and non-destructively detect the color of matcha.Combined with the Android mobile phone APP application and Bluetooth wireless communication technology to remotely control spectrometer equipment,which can greatly improve the practicability and operability of the system.The hardware system includes visible spectrometer,tungsten halogen lamp,Y-type fiber,sample cup,Bluetooth module and Android mobile phone.The visible light spectrometer model is SD1220,the effective wavelength range is 300~1000nm,the light source is halogen tungsten lamp,Y-type optical fiber is quartz material,the sample cup is 1.5ml centrifuge tube,the Bluetooth module model is HC-06 and Android phone model is vivo X6.The software includes Bluetooth,data processing and result display.The Bluetooth realizes searching,connection and communication of spectrometer.The others realize the real-time display of spectral data processing,L*a*b* color value.The research shows that the equipment can quickly detect the color of matcha.3.Study on rapid detection method of inner components of matcha using the near infrared spectroscopy technology.In this study,the inner components of matcha were detected by near infrared spectroscopy(NIR),and the quality grade of matcha was evaluated according to the ratio of tea polyphenols and amino acid.In this study,the contents of tea polyphenols and amino acids,two important internal components of matcha,were taken as the target.Collect diffuse reflectance spectra of matcha samples and then the standard chemical method was used to detect tea polyphenols and amino acid in each sample as the reference value for modeling.The spectrum is pretreated by SNV and then combined with full spectrum interval(Si),genetic algorithm(GA),competitive adaptive weighted random sampling algorithm(CARS)and random frog algorithm(RF)of four kinds of variable screening method established PLS quantitative prediction model of matcha polyphenols and amino acid.The comparison shows that the RF-PLS model has the best predictive effect and stability.Tea polyphenols detection model: Rp=0.8825,RMSEP=0.85%.Amino acid detection model: Rp=0.9662,RMSEP=0.14%.Finally,60 different grades of matcha samples are used to verify the feasibility of this method,a better discrimination effect was obtained.The results prove that using near infrared spectroscopy technology to rapidly and non-destructively detect the quality grade of matcha is feasible.4.The design of rapid non-destructive equipment for detecting the internal components of matcha.According to the characteristics of matcha processing technology,a rapid non-destructive testing device that mainly includes hardware systems and corresponding software systems was developed to quickly and non-destructively detect the internal components of matcha.Combined with the Android mobile phone APP application and BLE wireless communication technology to remotely control spectrometer equipment,which can greatly improve the practicability and operability of the equipment Hardware system mainly includes integrated near infrared spectrometer,source,sample cup and Android phone.Integrated near infrared spectrometer includes near infrared spectrum module,light source module and BLE module.Near infrared spectrum module is DLP NIRScan EVM spectrometer,whose wavelength range is 900 ~ 1700 nm,two small halogen tungsten lamps as light source module,BLE module model is CC2564 MODN.The power supply is two cores,the specification is 5V~1A,the capacity is 5000 mA.The sample cup is a cuboid quartz cuvette,android mobile phone model is m3 note.The software includes BLE,data processing and result display.The BLE realizes searching,connection and communication of spectrometer.The others realize the spectral data preprocessing,the internal component content calculation and the model prediction result display.The research shows that the equipment can quickly detect the inner components of matcha. |
PDF Full Text Request