| Jinhua ham is the famous traditional dry-cured ham in China. It is not only well-known at home, but also has a good fame overseas. However, presently the ham quality is mainly evaluated by a human taste panel. This method has many deficiencies, the organic evaluating has strong subjectivity, weak repeatability, and may vary by different factors, such as the physiological, mental reasons and experiences of human. As people are now paying more and more attention to food safety,it also proposes a higher request for Jinhua ham quality evaluating. Meanwhile, with the expansion of ham production and industrialization scale, the traditional method can no longer adapt to the modern requirements. So more scientific, reliable, objective and standard rules are urgent and important.This research focused on three different kinds of Jinhua hams, used the electronic nose (E-nose) technology, hyperspectral imaging technology and gas chromatography-mass spectrometry (GC-MS) technology to identify and classify them. In this dissertation, the main contents included:(1) Using E-nose to identify and classify different kinds of Jinhua hams. The data gained from E-nose were analyzed by Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA), the discriminative models were established by Partial Least Squares (PLS). (2) Using hyperspectral imaging technology to identify and classify different kinds of Jinhua hams. The data gained from hyperspectral imaging were analyzed by PCA and the discriminative models were established by PLS. (3) Using GC-MS to analyze the aroma components of different kinds of Jinhua hams, then comparing them with the results of E-nose and hyperspectral imaging.The research objective is to establish a reliable, rapid and nondestructive detection system for Jinhua hams, to provide new scientific methods and basis for Ham industry and promote the development of Jinhua ham industry.The main results and conclusions of this study included:1. Using the E-nose to detect different kinds of Jinhua hams. When analyzed by PCA,10 data of the 80th second gained at the stability phase of every grade were chosen. The variance of PCI was 74%, the variance of PC2 was 24%, the total variance was 98%. It meant that using PCA, E-nose can identify the top grade, the first grade and the second grade Jinhua hams. When it came to LDA,30 data of the 78-80 seconds gained close to the stability phase of every grade were chosen to be analyzed. The results of LDA showed that three kinds of Jinhua hams were well identified. And there were some cross between the first grade and the second grade hams. But generally speaking, LDA can identify different kinds of Jinhua hams.10 data of the 80th second were also analyzed by PLS to establish discriminative models. The total recognition rate of training set and checking set were 95% and 88.33%. And that the erroneous judgement was mainly happened between the top grade and first grade hams, while the probability of erroneous judgement with the second grade ham is low.2. Using hyperspectral imaging technology to detect different kinds of Jinhua hams. The band region of 1000 nm～1600 nm was selected and the results showed that the images at 1001 nm and 1328 nm can mostly characterize the original information of samples. So the two wavelengths were selected as the characteristic wavelengths. 256 bands were gained by Region of Interest (ROI), then chose 160 continuous bands and gained their average spectrum values. The data were analyzed by PCA, and the results showed that hyperspectral imaging can identify different kinds of Jinhua hams well. Establishing discriminative models by PLS, the results showed that the total recognition rate of training set and checking set were 96.19% and 89.52%. And that the erroneous judgement was mainly happened between the top grade and first grade hams, while the probability of erroneous judgement with the second grade ham is low.3. Using GC-MS with solid-phase micro-extraction (SPME) to detect different kinds of Jinhua hams. Choosing the extraction results of 100μm PDMS and the effective peak time was 19-34 minutes. Comparing the chromatogram with MS library and the results showed that Benzene,2,4-diisocyanato-l-methyl-and 2H-Benzimidazol-2-one,1,3-dihydro-5-methyl-were detected in all three kinds of Jinhua hams. The two components had a high level in all three kinds, but there were no obvious difference among them. So their Contribution rate for identifying hams is small. But they are important for the total aroma of Jinhua ham. Butylated Hydroxytoluene and Tetradecanal were also detected in all three kinds of Jinhua hams and the different among them is big. So it could be believed that they played an important role in the different aroma of Jinhua hams. Moreover, Oxalic acid, neopentyl pentyl ester was only detected in the top grade ham,2,2,4-Trimethyl-1,3-pentanediol diisobutyrate was only detected in the first grade ham,4-Hexen-3-ol and Phytol were only detected in the second grade ham. Although the content of these components was not so much, they also played an important role in the different aroma of Jinhua hams. |
PDF Full Text Request