| To improve the efficiency of qualities detection for germinated brown rice(GBR) in the production practice, the change for quality index of germinated brown rice were investigated under different microwave drying conditions, and non-destructive detection method for its quality was established based on near infrared spectrum(NIS) technology, and the comprehensive quality of germinated brown rice was evaluated, as well as quality evaluation standards were initially formulated. The main conclusions in this research are as follows.By means of the single factor experiment method, the response characteristics of the near infrared spectral for germinated brown rice samples were determined on different scanning temperature and scanning times, and the scanning temperature and scanning times were deceded,as well as the detection parameters of NIS were optimized. The result suggested that suitable NIS scanning conditions could improve the detection performance of transmitted spectrum. The average of relative standard deviation in NIS absorbance was minimum at the temperature of 25℃,and the average of relative standard deviation was minimum when the same GBR samples were scanned at three times.The spectrum of 105 samples for GBR were collected from measurement of the near infrared spectrum analyser, and the characteristics absorption wavelengths of GBR were determined. The spectrum and chemical values of quality indicators of germinated brown rice were measured according to the national standard method. The changes GBR quality were analysed during the microwave drying process. The result showed that obvious composition differences were distinguished between samples by NIS method. NIS absorbsance of quality characteristics for GBR were presented at the five different wavelengths. Comparing with raw GBR(no drying),microwave drying processing improved quality indexes of the GBR in terms of starch, moisture,protein, lightness value, redness, and yellowness value The change of GABA content in GBR showed increasing followed by decreasing trends.The abnormal concentration of samples could be identifed and eliminated by using significance test method based on predicting concentration residual of F test. Mathematical models of characteristic absorbent value of NIS and quality index including physical and chemical values of GBR were built by using MLR, PCR and PLSR. A optimal mathematical model was determined.by using Center Composite design Experiment method with four-factor and five-level. The verified results showed that the accuracy of the models was improved by deleting abnormal samples. Theoptimal mathematic model about GBR in terms of starch, moisture, protein, GABA content,lightness value, redness value, yellowness value could be established by MLR, and the determination coefficient were 0.9014, 0.9005, 0.9448, 0.9430, 0.9498, 0.9007, 0.9404,respectively, and prediction standard deviation were 2.9279, 1.4950, 1.0374, 1.0038, 1.0950,1.4950, 0.2950. The correlation coefficient of GBR between predictvalues and real values were0.9519, 0.9644, 0.9532, 0.9037,0.9490, 0.9162, 0.9164, respectively, and prediction standard deviation were 2.6882, 1.3674, 1.0427, 0.5988, 0.9962, 0.9528, 0.4827, respectively.Quality indexes of GBR in terms of starch, moisture, protein and gamma aminobutyric acid,lightness value, yellowness value, redness value were selected to synthetically assess and grade the quality of germinated brown rice. The weight of each index and the comprehensive score of each sample were calculated by using Entropy Weight Coefficient Method. The grading method proposed in this paper was feasible by validation. The samples could be obtained based on Center Composite design Experiment method with four-factor and five-level. Effects of different microwave drying condictions on quality of germinated brown rice were analysed, as well as quality standard of germinated brown rice was developed to evaluated and test the GBR quality state. |
PDF Full Text Request