Congou Objective Evaluation Of Quality

All along, the quality control and quality evaluation congou more subjective sensory evaluation mainly because of individual differences, subjective sensory evaluation influenced by various factors and constraints, it is difficult to work on the quality of black tea quantitative evaluation and analysis. In this study, a random sample of 2013 congou tea samples, near-infrared spectroscopy, computer machine vision technology, the color spectrum technology to work in an objective assessment of the quality of black tea. Investigate objectively evaluate technical and sensory evaluation of consistency, to establish the quality of tea objective of the rapid evaluation tools for the analysis of the identifying characteristics of the new ideas, to lay a new foundation for the quality evaluation of the quality of tea. The results will now be reported as follows:1. Based on Near Infrared Spectroscopy (NIRS) and objective evaluation of the work of black tea quality researchUsing near infrared spectroscopy scanning congou tea samples atlas combines sensory evaluation score result, the content contains 12 kinds of substance use OPUS7.0 quantitative analysis software to build prediction models, predictive better.Near-infrared spectroscopy combined with the results of sensory evaluation scores established quantitative models in addition to the aroma, Re and Rp models were more than 85%, the effect of scaling all models better, something that can be used to predict the quality of tea, one of the representatives congou comprehensive quality tea samples sensory quality score highest prediction accuracy, a single factor in the soup, taste and Securinega higher, Re were more than 86%, Rp were more than 92%. According to the shape of each model, soup, aroma, taste, the bottom and out of Rp 86.22%,97.02%,48.04%,99.30%,96.86%,92.34%;the corresponding RMSEP were 0.0168,0.0203,0.0357,0.0266,0.0251,0.105. In addition to the aroma, the prediction accuracy than higher.Among them, the prediction accuracy is not excellent aroma, Rp value of only 48.04%, indicating that the model is not suitable for near-infrared congou aroma factor prediction remains to be further studied. Score, soup, taste and Securinega model prediction accuracy are excellent, Rp values were above 92%. Therefore, the use of near-infrared modeling described the effort to get the comprehensive quality black tea and soup senses, taste, excellent performance prediction Securinega factor for congou tea sensory evaluation of comprehensive quality prediction.With the present experiment HPLC liquid chromatography (HPLC), this method is advantageous in that the contents can measure the seven kinds of substances (caffeine, C, EC, EGC, ECG, EGCG, TFs) in the same liquid phase conditions, such that efficient liquid chromatography (HPLC) determination of tea contains components more convenient, fast and accurate.Near-infrared spectroscopy combined with 12 kinds of quantitative forecasting model component content material containing quantitative model established, the correlation coefficient Rc and Rp each component models are up more than 90 percent, according to the caffeine, C, EC, EGC, ECG, EGCG, TFs, thearubigins, Chahe hormone, polyphenols, amino acid sequence Rp values were 99.88%,97.12%,99.39%,99.66%,99.25%,99.18%,99.73%,92.89%,91.41%,97.41%,98.46%,97.58% ;RMSEP0.0773,0.00806,0.0745,0.00759,0.0756,0.0765,0.0362,0.0762,0.0752,0.0148,0.00915,0.01 68. The results showed that the correlation coefficient prediction Rp are above 90%, RMSEP value is relatively small, less than 0.1, accurate prediction performance model, the selected component congou 12 have reached a quantitative analysis of the requirements, which can be obtained known test the model can accurately predict the content of the work-related components of black tea leaves, indicating near-infrared spectroscopy techniques can be applied to something black tea quality attributes quantitative analysis. Congou Tea predict molecular structure of each component is relatively simple, relatively simple functional groups contained in the near infrared absorption frequency combined with strong beneficial information extraction near-infrared spectroscopy, the ideal model for prediction.Thus informed, congou using near infrared spectroscopy combined with tea sensory evaluation score results, tea contains 12 kinds of component content material build predictive models, except the aroma index score, prediction performance better. So Near Infrared Spectroscopy to quickly and accurately assess the quality of tea, black tea is something objective evaluation techniques to provide a broad prospect.2. The objective evaluation of the work of black tea quality computer-based machine vision technology researchMachine vision technology using computer images congou intake, extract color feature parameters, combined with the quality of tea grading method DPS7.05 software BP neural network, the main conclusions and recommendations are as follows:As can be seen from the color characteristic parameters statistics, the average premium congou tone was significantly higher than the average of the other two levels of tea, black tea thus explain something more significant premium over the secondary level color parameters. Although one, two tea samples of H, S, I mean closer parameters, but clear differences between the standard deviation 5, a congou standard deviation is greater than two.For the shortcomings of traditional identification methods, the use of artificial neural network approach something black tea quality grading. There is a selection of hidden layer BP network; input can be different depending on the selected parameters and the tea set, the test input nodes 18 (dry tea, tea, the bottom color characteristic value); output node 3, on behalf of tea 3 grades (grade, one, two).BP neural network congou rank correlation coefficient of determination of tea reached 0.9978 forecasting model to predict superior performance, so you can come take advantage of this neural network forecasting model to predict congou grade identification.Use of software design DPS7.05 BP network genetic algorithm extracted 2013 60 congou training and testing, the average recognition accuracy was 93.3%, better grading. Proof computer machine vision congou quality grading feasible for congou machine vision technology combined with computer online quality grading objective evaluation provides a broad prospect.3. The objective evaluation of the quality of work based on the color of black tea spectroscopySpectroscopy combined with the use of color sensory evaluation score results for 2013 randomly selected 30 congou objective evaluation of the quality of tea samples.Congou tea sensory evaluation score results and tea color values are highly correlated, and L which was a significant negative correlation, the correlation coefficient was 0.8030; and a, b, derived from the value of b/a, Cab, Sab,ΔE showed a significant positive correlation, the correlation coefficients were 0.9954,0.9926,0.8957,0.9930,0.9758,0.9787; coefficient of determination were 0.6448,0.9908,0.9853,0.8023,0.9860,0.9522,0.9579.Theaflavins TFs and color values L was a significant positive correlation, the correlation coefficient was 0.54 **, and a, b, b/a, Cab, Sab, AE are a significant negative correlation between the order of the correlation coefficient -0.70 **. - 0.69 **. - 0.58 **. - 0.69 **. - 0.67 **. -0.68 **; thearubigins TR and L was significantly negatively correlated, the correlation coefficient was -0.61 ** with other color values are a significant positive correlation with the a, b, b / a, Cab, Sab, AE correlation coefficients were 0.82 **, 0.80 **, 0.68 ** ,0.80 ** ,0.76 **, 0.79 **; Chahe hormone TB and L was significantly positively correlated with a correlation coefficient of 0.67 **; and other color values showed a significant negative correlation was significant, with a, b, b / a, Cab, Sab, A E correlation coefficients were -0.81 **, - 0.79 ** - 0.68 * 0.80 **, 0.79 ** 0.78 **From the above that there is something between black tea pigment color value and relevance for: L values greater, indicating that more and Chahe theaflavins pigment content, the lower thearubigins content; the larger the a, b values show thearubigins content, the lower theaflavins and Chahe pigment content.Catechins EGC and thearubigins TR and sensory evaluation scores were significantly positively correlated, the correlation coefficients were 0.76 ** ,0.82 **; the other 10 indicators caffeine, C, EC, ECG, EGCG, TFs, TB, polyphenols, amino acids, water and reviewers are scores were significantly negatively correlated, and the correlation coefficient were -0.58 **,- 0.75 **, - 0.79~ *, - 0.71 **,- 0.69 **, - 0.69 **, - 0.80 ** ,- 0.70 ** ,- 0.80 **, - 0.69 **The results showed that when the time black tea catechins EGC and thearubigins TR content is relatively high, higher review scores, better quality tea time; the other 10 indicators caffeine, C, EC, ECG, EGCG, TFs, TB, relatively low polyphenols, amino acids, water, high review scores, better quality tea.Thus informed, the use of color congou measured parameters, and Lab color color color table analysis, fast, accurate, and objective evaluation of the quality of work of black tea, black tea is something objectively evaluate technical foundation.In summary, based on near-infrared spectroscopy technology, computer technology and the color spectrum of machine vision technology for black tea quality work to overcome the limitations imposed by various subjective factors and the impact of a higher degree of objective evaluation method is accurate, rapid, objective evaluation of the prospects broad.