Study On Portable Detection Of Ripeness Of Orange Based On Spectral Index

China is one of the rich tangerine country and export volume increased year by year,the tangerine is famous for its sweet.It has the advantage of nutrient rich and can prevent many diseases,in recent years,the tangerine was exported to the Southeast Asia,Middle East,European Union and other regions.However,judging the maturity of tangerine was by the farmers who with the rich experience currently,as the different of individual experience could lead to the tangerine which entered the market had different maturity and then effected the transportation,storage and sales,this situation would reduce its market competitiveness and export.In recent years,the methods to detecting fruit maturity include acoustic detection,NIR nondestructive testing,hyperspectral detection,electronic nose and so on.However,the detection system of these methods is complex,huge volume,expensive and inconvenient to operate.Therefore,it is important to study the internal quality of tangerine maturity and develop a simple system,convenient operation and low cost tangerine maturity detection device.The main contents and conclusions of the study are as follows:(1)Comparative analysis the index of tangerine maturity based on the physicochemical characteristics.The study take the tangerine in the September 2016-11 month as the research object,using the method of statistical analysis the tangerine weight,transverse diameter,longitudinal diameter,chlorophyll,soluble solids,pH,color,solid acid ratio,ratio of chlorophyll and soluble solids,chlorophyll and the ratio that chlorophyll and the acidity of solid acid and the ratio of the mean value and deviation,the results show that the three parameters including ratio of chlorophyll,the ratio of chlorophyll and soluble solids,the ratio of chlorophyll and soluble solids acid ratio can as the evaluate index for tangerine maturity.(2)Tangerine visible/near infrared spectral analysis.Through analysis the characteristics of average spectrum of 6 harvest time with the changes of harvest period,the results showed that the formation of visible/near infrared spectra peaks had the relationships with the absorption band of chlorophyll,according to it gained four characteristic wavelengths were649nm,672nm,724nm,1100nm respectively.Further analysis the coefficient of variation of the spectrum,the results showed that the coefficient variation of tangerine Vis/NIR spectra in different harvest time had the maximum in the 670nm wavelength,at the same time,it is the absorption peak of chlorophyll.(3)The comparison analysis of evaluation index of tangerine maturity based on the spectral index.Using the four characteristic wavelengths obtained before combined with the method of spectral index construction to constructed 6 kinds of the evaluation index of tangerine maturity,taking the correlation coefficient as the evaluate index to evaluate index of the 6 kinds of tangerine maturity,the results showed the highest correlation coefficient is0.9916,so using it as the best tangerine maturity evaluation index.(4)The evaluate model establishment and verification of tangerine maturity based on the visible/near-infrared spectroscopy.Combined the evaluate index of tangerine maturity and the best evaluation index of tangerine maturity to establish multiple linear regression model,the results showed that take chlorophyll as the model evaluation index of tangerine maturity could obtain the best model effect and the correlation coefficient of modeling set and the prediction correlation coefficient were 0.98 and 0.96 respectively,thee root mean square error of modeling and prediction were 0.49 and 0.59,modeling and forecasting deviation is-6.1×10~-88 and-0.014.the multiple linear regression equations of calibration set and the prediction set were y=0.95x+0.016 and y=0.91x+0.032.(5)Microprocessor based maturity detection system.The system is composed of 650nm and 724nm wavelength lamp,STC89C52 micro-controller,silicon photocell,AD7705 module and PC machine.Take the tangerine of 9-11month in 2017 as the research object,using the system to measure the output value of voltage of tangerine in different harvest period of two different wavelength which lighted by silicon photocell.Finally,combined the ratio of measured voltage and chlorophyll to establish multiple linear regression equation.The results show that the correlation coefficient of modeling set and prediction set was 0.77 and 0.75respectively and the root mean square error of modeling and the prediction set was 3.7 and 3.8respectively.The multivariate linear regression equation of modeling set and prediction set is y=0.768x+1.865 and y=0.775x+2.432 respectively.