1998 Volume 60 Issue 3 Pages 117-127
The near infrared reflectance system for on-line quality inspection of fruits was developed and examined to predict individual sugars and malic acid concentrations of cylindrical samples of the flesh of Golden Delicious apples and intact mangoes. Two kinds of multiple linear regression methods were examined to calibrate the log (1/R) values in wavelengths from 1400 to 2000nm with individual sugars and malic acid concentration determined by a chromatography procedure. A set of calibration equations developed by using stepwise multiple linear regression was found to be applicable for predicting the individual sugar and malic acid of the tested fruits. The correlation coefficients of the relationships between actual values and predicted values were more than 0.9 with standard error predictions less than 0.44%.