Abstract: Near infrared spectroscopy models for nondestructive detection of sugar content and acidity of Nanguo pears were established based on the spectrum recorded by Purespect near infrared instrument and the content of soluble solids (SSC),reducing sugar (RSC),and titratable acidity (TA) and pH value determined by chemical methods. Based on the comparison results of correlation coefficient of calibration (Rc),correlation coefficient of prediction (Rp),root mean standard error of calibration (RMSEC),root mean standard error of prediction (RMSEP) and sum of residual error,the best parameters which were used for building up NIR models were confirmed. Results showed that the RC and RP value of SSC are higher than those of RSC,and the residual error is better distributed Therefore,SSC is the optimum parameters. NIR models for detection of sugar. RC value of pH is higher than that of TA,RP of both TA and pH are above 0.9,and the residual error sum of TA is better than pH. Considering the purpose of practice application,TA is the optimum parameter for NIR models for detection of acid.

Key words: Nanguo pear, NIR spectrum, sugar content, acidity