Systematic non-destructive test of the apple internal quality based on MATLAB software

The objective of this study was through linking nearly infrared equipment with MATLAB software to realize online non- destructive testing the fruit internal quality based on integration of infrared light scanning, data processing, prediction model and model evaluation of apple sugar and titratable acidity. Through writing and debugging the MATLAB modularized program of the near infrared spectra pretreatment, partial least squares ( PLS) regression analysis, quantitative prediction model, prediction results and model accuracy test to get the quantitative forecasting and model evaluation of apple sugar and titratable acidity.The results showed that the correlation coefficient ( R) between predicted values and real values, Standard error of correction ( SEC) , Standard error of prediction ( SEP) , and the standard deviation ( SD) of apple sugar ( titratable acidity) builded the prediction models were 0.9528 ( 0.8786) , 0.4788 ( 0.0215) , 0.3170 ( 0.0128) , and 1.4111 ( 0.0390) , respectively.It had been demonstrated that rapid quantitative forecasting the content of apple sugar and titratable acidity using Near Infrared reflectance spectroscopy based on MATLAB 7.1 were feasibility, and the model had a good prediction accuracy.The method of quantitative online non- destructive prediction the content of apple sugar and titratable acidity will have a important significance to develop internal quality automation evaluation software of fruit and classification software.