CN106501212A - Based on the method that the ripe rear quality of beef is roasted in the information prediction of raw meat near infrared spectrum - Google Patents

Abstract

The present invention relates to the technical field of quality detection of meat products, in particular, to a method for predicting the quality of grilled beef after cooking by using the near-infrared spectrum information of raw meat. The method includes the following steps: a). Obtaining raw beef for barbecue to be detected The near-infrared spectrum of the sample; b). Through the pre-built near-infrared spectrum data model, the near-infrared spectrum obtained in step a) predicts the cooked quality of the raw beef sample for barbecue to be detected; wherein, the The near-infrared spectrum data model includes the corresponding relationship between near-infrared spectrum and parameters related to ripening quality.

Description

Method for predicting cooked quality of grilled beef based on near-infrared spectral information of raw meat

technical field

The invention relates to the technical field of quality detection of meat products, in particular to a method for predicting the cooked quality of grilled beef based on near-infrared spectrum information of raw meat and its application.

Background technique

With the acceleration of the pace of life in modern society and the continuous integration of Chinese and Western food cultures, barbecue beef is becoming more and more popular among young consumers with its unique flavor and convenient and fast cooking methods.

Raw materials are one of the main factors affecting the consumption value of barbecue beef. The beef roasted by different parts and different quality raw materials often have great differences, especially in the Chinese market. Raw materials of different quality put forward higher requirements for the quality evaluation and grading of barbecue beef. Traditional beef quality testing and grading focus more on carcass evaluation during bulk transactions, and are difficult to apply to specific barbecue products, let alone predict the quality of cooked raw materials, and mostly use destructive methods such as chemical analysis and instrument testing Mainly, not only the process is complicated, time-consuming and labor-intensive, but also causes a lot of raw material pollution and waste. Therefore, in order to promote the development of the beef industry, it is necessary to establish a set of rapid and effective detection methods and grading methods for the quality characteristics and consumption needs of barbecue beef, so as to realize the rapid detection of raw meat for barbecue.

Near infrared (NIR) is an electromagnetic wave between visible light (VIS) and mid-infrared light (IR). The American Society for Testing and Materials (ASTM) defines it as a spectral region with a wavelength of 780-2526nm. A modern analytical technique developed since the 1970s, which is widely used in various fields. Its sample processing method is simple, and multiple indicators of meat can be evaluated at the same time. At present, it is mainly used for the identification of origin and variety , quality evaluation and safety testing. The near-infrared spectroscopy technology has the characteristics of fast, accurate, and pollution-free, which is of great help to improve work efficiency, reduce raw material loss, and reduce labor intensity. Some studies have shown that near-infrared spectroscopy has a good predictive effect in the analysis and prediction of meat and meat products such as pork, beef, mutton, chicken, ham, and cured meat. The application of grading is still rarely reported, and in this field, the method of detecting raw beef for barbecue to predict its cooked quality is still blank.

In view of this, the present invention is proposed.

Contents of the invention

The first object of the present invention is to provide a method for predicting the cooked quality of barbecued beef based on the near-infrared spectrum information of raw meat. The method can predict the cooked quality of raw raw beef for barbecue through rapid detection.

The second object of the present invention is to provide an application of the method for predicting the cooked quality of roast beef based on the near-infrared spectrum information of raw meat in rapid grading of raw beef for barbecue.

In order to realize the above-mentioned purpose of the present invention, special adopt following technical scheme:

A method for predicting the cooked quality of grilled beef based on near-infrared spectrum information of raw meat, comprising the steps of:

a). Obtain the near-infrared spectrum of the raw beef sample for barbecue to be tested;

b). Through the pre-built near-infrared spectrum data model, the cooked quality of the raw beef sample for barbecue is predicted by the near-infrared spectrum obtained in step a);

Wherein, the near-infrared spectrum data model includes the corresponding relationship between near-infrared spectrum and cooked quality-related parameters.

Preferably, the method for predicting the cooked quality of grilled beef based on the near-infrared spectrum information of raw meat as described above, the near-infrared spectrum data model is obtained by the following method:

b1). After establishing a large sample data of the near-infrared spectrum of raw beef for barbecue, it is tested for parameters related to the cooked quality;

b2). Establish a near-infrared spectrum data model according to the corresponding relationship between the large sample data of the infrared spectrum and the parameters related to the cooked quality.

Further preferably, in the method for predicting the cooked quality of roasted beef based on the near-infrared spectrum information of raw meat as described above, the parameter related to the cooked quality is a sensory score.

For the sensory evaluation method, refer to GB/T 22210-2008 "Standards for Sensory Evaluation of Meat and Meat Products" and the general principles and methods established by Zhao Lei et al. The method mentioned in .

The sensory evaluation is carried out on the roasted beef, and the evaluation criteria of the sensory evaluation will be specifically described in an embodiment of the present invention.

Preferably, in the method for predicting the cooked quality of roasted beef based on the near-infrared spectrum information of raw meat as described above, the state of the raw beef for roasting is minced meat.

Preferably, the method for predicting the cooked quality of roasted beef based on the near-infrared spectrum information of raw meat as described above, the raw beef for roasting is taken from two parts: the outer spine and the cucumber strips.

It is worth noting that, in the method for predicting the cooked quality of roasted beef based on the near-infrared spectrum information of raw meat provided by the present invention, when establishing the near-infrared spectrum data model, establishing the relationship between the near-infrared spectrum and the sensory score is the most direct and concise method , but in addition, it can also be used as a sensory evaluation by measuring the main chemical components (protein, fat, water content) and quality indicators (WBSF, WHC, CL, L*, a*, b* value) of beef raw materials auxiliary indicators.

Preferably, in the method for predicting the cooked quality of grilled beef based on the near-infrared spectrum information of raw meat as described above, the establishment method of the near-infrared spectrum data model is the discriminant partial least squares method.

Preferably, the above-mentioned method for predicting the cooked quality of grilled beef based on the near-infrared spectrum information of raw meat, in the method, the acquired near-infrared spectrum is a spectrum within the range of 1000-2500nm.

Preferably, the method for predicting the cooked quality of roasted beef based on the near-infrared spectrum information of raw meat as described above, in the method, the method for obtaining the near-infrared spectrum is to scan and acquire with a near-infrared spectrometer, and the average number of scans is 30 times .

Preferably, in the method for predicting the cooked quality of grilled beef based on the near-infrared spectrum information of raw meat as described above, the near-infrared spectrometer is a SupNIR-1550 portable near-infrared spectrometer.

The application of the method for predicting the cooked quality of roast beef based on the near-infrared spectrum information of raw meat as described above in the rapid grading of raw beef for barbecue.

Compared with prior art, the beneficial effect of the present invention is:

The present invention is very necessary for large-scale meat production enterprises to explore the accurate prediction of the quality of roast raw beef after cooking. The invention explores the conditions for the domestic portable near-infrared spectroscopy instrument to be used for spectral scanning in the process of measuring raw beef, and provides a theoretical basis for exploring the best technical conditions for predicting the quality of roasted beef after cooking. Using the method provided by the invention to predict the quality of cooked beef can realize the effect of saving time and capital costs of enterprises, and the prediction method provided by the invention is fast, simple and repeatable.

Description of drawings

In order to more clearly illustrate the specific implementation of the present invention or the technical solutions in the prior art, the following will briefly introduce the accompanying drawings that need to be used in the specific implementation or description of the prior art. Obviously, the accompanying drawings in the following description The drawings show some implementations of the present invention, and those skilled in the art can obtain other drawings based on these drawings without any creative work.

Fig. 1 is the sensory score distribution chart of the sample in the embodiment of the present invention;

Fig. 2 is the near-infrared average spectrogram of sample in the embodiment of the present invention;

Fig. 3 is the near-infrared average spectrogram of the sample in the minced meat and meat block state after the first derivative treatment in the embodiment of the present invention;

Fig. 4 is a comparison distribution diagram of actual classification and predicted value of raw beef prediction set meat samples in the minced meat state in the embodiment of the present invention.

detailed description

Embodiments of the present invention will be described in detail below in conjunction with examples, but those skilled in the art will understand that the following examples are only for illustrating the present invention, and should not be considered as limiting the scope of the present invention. Those who do not indicate the specific conditions in the examples are carried out according to the conventional conditions or the conditions suggested by the manufacturer. The reagents or instruments used were not indicated by the manufacturer, and they were all conventional products that could be purchased from the market.

Example

Establishment and Application of Near Infrared Spectroscopy Data Model

1. Materials and Methods

1.1. Materials and reagents

Outer loin (longissimus dorsi) and small cucumber strips (semitendinosus) were respectively selected from 20 and 60 Simmental hybrid cattle fattened by Beijing Yuxiangyuan Group and Ningxia Xiahua Muslim Meat Food Company, and were slaughtered at 0- Acid discharge for 48h at 4°C. After the beef is marinated, the loin and gherkin strips are removed on the spot.

1.2. Instruments and equipment

SupNIR-1550 portable near-infrared spectrometer: Hangzhou Concentrating Technology Co., Ltd.

1.3. Test method

1.3.1. Spectrum acquisition

The beef is cut into a complete sample with a thickness of 3cm and a size of about 3×6×6cm ³ ; the minced meat is a homogeneous sample obtained after mincing under the condition of 2.0×10 ³ rpm for 15s, and the grating dispersion is applied under the dark background The spectrum is collected by a portable near-infrared spectrometer with a wavelength range of 1000-2500nm and a resolution of 1nm. The average number of scans of the instrument is 30 times, and the average spectrum is obtained by scanning the sample 3 times.

1.3.2. Baking

Cut the measured meat piece into 2.54cm thick, weigh it, insert the inductive coupling thermometer into the center of the meat piece, roast at about 160°C until the center temperature of the meat piece reaches 70°C, and store the baked meat piece in aluminum foil Cool to room temperature and set aside.

1.3.3. Sensory evaluation

Choose 10 people who are engaged in food quality and safety to form a sensory evaluation team, and train the evaluation personnel on sensory quality indicators such as roasted beef color, tenderness, flavor, and juiciness. They are required to understand the meaning and subjective feelings of the comments used. The evaluation method adopts a 7-point system for the evaluation of various quality indicators of barbecue, that is, 1 to 7 points, see Table 1. The evaluation results are calculated according to the evaluation model of barbecue beef consumer preference and satisfaction made by our research group in the previous period. The final weighted score is calculated. Calculated as follows:

Sensory score = 0.167 x color + 0.237 x tenderness + 0.206 x flavor + 0.170 x juiciness.

Table 1 Sensory evaluation table of roasted beef quality

Table 1Roast beef Sensory Evaluation

1.4. Data analysis

The format of the collected spectral data was converted, and the prediction model was established using The Unscrambler (version 9.8, CAMO), and the qualitative discrimination of the sensory grade was completed by the discriminant partial least squares method (PLS-DA). The correlation coefficients R ² _c and R ² p between the measured value and the predicted value of the calibration set and prediction set samples, the root mean square error of internal cross-validation (RMSECV) and the root mean square error of prediction (RMSEP) were used as the evaluation model quality indicators.

2. Results and Analysis

2.1. Statistics of sample quality indicators

The 160 beef samples were used to establish a calibration set and a prediction set at a ratio of 3:1, and the sensory evaluation results are shown in Table 2. The accuracy of near-infrared in predicting samples is not only affected by the measurement accuracy of each index, but also by the amount of information contained in the spectrum and the variation range of reference data. When the data set is established, some eigenvalues (maximum value, minimum value, etc.) are included in the calibration set to ensure that the range of the calibration set includes the range of the prediction set, ensuring the applicability and reliability of the built model.

Table 2 Statistical table of beef sample quality indicators

Table 2quality traits statistics of beef samples

2.2. Barbecue beef sensory grade classification

The sensory score distribution of roasted beef raw materials is shown in Figure 1, where the highest score is 5.9968 and the lowest is 3.2037. According to the sensory score, the existing beef samples can be divided into three grades, that is, the first grade (5-6 points), the second grade (4-5 points) and the third grade (3-4 points).

2.3. Spectral analysis and preprocessing

2.3.1. Near-infrared average spectrum

The average spectra of the beef samples in the block and mince state are shown in Figure 2. It can be seen from the figure that there are strong absorption peaks at 1150nm, 1450nm and 1930nm, which are the first-order multiplier and combined frequency of the O-H bond, because most of the beef is water, with a content of more than 70%, and the absorption peak of the O-H bond very obvious. The first frequency doubling of C-H at 1600-1800nm, the second frequency doubling of 1100-1400nm and the first frequency doubling of N-H at 1400-1600nm, and the absorption peak of -OH at 2100nm can also be identified from the figure. Indicates changes in the content of organic matter such as protein and fat. Because protein and fat content are decisive factors affecting the quality of grilled beef, changes in these absorption peaks are especially important for prediction accuracy. It can be seen from the figure that the trends of the near-infrared spectra collected under different sample states are roughly the same, but in the range of 1000-1900 nm, the spectrum in the minced meat state has a larger fluctuation range and the absorption peak is more obvious.

2.3.2. Spectral preprocessing

The prediction accuracy of near-infrared spectroscopy is often interfered by factors unrelated to sample properties, such as ambient temperature, sample state, light scattering, and instrument response, which lead to baseline drift and poor repeatability of near-infrared spectroscopy. Therefore, it is necessary to analyze and preprocess the collected near-infrared spectral data to eliminate these adverse effects and improve the prediction ability. In this study, different methods such as smoothing, standard norm variate (SNV), multiplicative signal correction (MSC), and derivative processing (Derivative) were used to process near-infrared spectra to eliminate various high Frequency noise and baseline drift, improve repeatability and signal-to-noise ratio.

Figure 3 is the near-infrared spectrum after the first derivative processing. It can be seen from the figure that after the spectrum is processed by the first order derivative, the linear baseline drift is effectively reduced and the band characteristics are strengthened. In the collected spectral data, the O-H absorption peak at 1150nm, the C-H absorption peak at 1200-1400nm and the N-H absorption peak at 1400-1600nm are all more obvious. Among them, the absorption peaks of the near-infrared spectrum collected under intact meat pieces fluctuated more in the range of 1000-1100nm and 1400nm, and the absorption peaks in the minced meat state were more prominent in the ranges of 1200-1400nm and 1500-1900nm.

2.4. Qualitative discrimination of sensory grades of barbecued beef

2.4.1. Establishment and prediction of partial least squares discriminant model for samples of different grades

Table 3 shows the prediction results established by the discriminant partial least squares model (PLS-DA). It can be seen from the table that after optimization of preprocessing methods such as smoothing (Smooth), multivariate scattering correction (MSC), standard normalization distribution (SNV), and derivatives (Derivatives), the grade prediction results established in the minced meat state are significantly better in the state of meat. In the minced meat state, the spectrum can properly eliminate background interference after 3-point smoothing (Smooth-G(3)) processing, and achieve the best prediction effect. The correctness rates of the calibration set and verification set are 95.00% and 93.33%, respectively. The correct rate of the prediction set was 85.00%, and the R ² (coefficient of determination) and the standard deviation of cross-validation (RMSECV) were 0.81 and 0.28, respectively (Fig. 4). It can realize the prediction of unknown samples.

Table 3 Prediction results of different spectral preprocessing methods

Table 3predict results for PLS-DA models with different spectral pretreatments

Note: None: without any preprocessing; Smooth-G(N): N-point smoothing processing; MSC: multivariate scattering correction processing; SNV: standard normalized distribution processing; S-G(N)+1D: N-point first derivative Smoothing; De-trending: detrending; S-G(3)+SNV: 3-point smoothing plus standard normalization; S-G(3)+1D+SNV: first-order derivative 3-point smoothing plus standard normalization .

3 Discussion

Sensory is one of the most direct and effective means to evaluate the quality of grilled beef, but the prediction of sensory indicators by near-infrared has been difficult to obtain satisfactory results. On the one hand, it is because people have different subjective feelings about the color, tenderness, flavor, juiciness and other indicators of beef, and the judgment results will be quite different. At the same time, it is difficult to make accurate quantitative judgments like objective indicators. . On the other hand, different sensory indicators often get a small range of variation after weighted calculation scores, which reduces the accuracy of near-infrared model prediction. In addition, the samples at the time of spectral scanning and the samples scored by the sensory evaluation panel are not exactly the same, and the changes in quality and the influence of cooking process will reduce the prediction accuracy.

In order to improve the prediction accuracy, this study selected two parts of meat, the outer ridge and the small cucumber strips, as barbecue raw materials to expand the range of variation of the scores. At the same time, the sensory scores of all samples were divided into three grades in the prediction of sensory results. The problem of small index variability is avoided, and finally a better prediction effect is achieved through near-infrared modeling and prediction.

In this study, beef tenderloin and gherkin strips were selected as samples for barbecue, and a portable near-infrared spectrometer was used to collect spectra of fresh samples in the factory operating environment in the range of 1000-2500 nm, and sensory evaluation was performed after roasting. Using 160 samples as the modeling set, the partial least squares discriminant method (PLS-DA) was used to classify and predict the sensory score results. Finally, the correctness rates of the calibration set and verification set established in the minced meat state were 95.00% and 95.00%, respectively. 93.33%, the correct rate of prediction set discrimination is 85.00%, R ² and RMSECV are 0.81 and 0.28 respectively. The results show that the near-infrared technology can be successfully used in the rapid quality detection and grade discrimination of raw beef meat for barbecue in the factory environment.

In summary, the present invention establishes a near-infrared spectrum data model by collecting the near-infrared spectrum information of raw meat, and can conveniently obtain the sensory quality of grilled meat through this model, thereby saving time and money for related manufacturers and enterprises. Reduce capital costs, avoid unnecessary waste of low-quality beef, and provide cost-saving, scientific and reliable basis for determining the barbecue method of raw meat and pricing barbecue meat.

It should be noted that at last: above each embodiment is only in order to illustrate technical scheme of the present invention, and is not intended to limit; Although the present invention has been described in detail with reference to foregoing each embodiment, those of ordinary skill in the art should understand that: It is still possible to modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the technical solutions of the various embodiments of the present invention. range.

Claims (10)

1. A method for predicting the cooked quality of roasted beef based on raw meat near-infrared spectrum information, characterized in that, comprising the steps: a). Obtain the near-infrared spectrum of the raw beef sample for barbecue to be tested; b). Through the pre-built near-infrared spectrum data model, the cooked quality of the raw beef sample for barbecue is predicted by the near-infrared spectrum obtained in step a); Wherein, the near-infrared spectrum data model includes the corresponding relationship between near-infrared spectrum and cooked quality-related parameters. 2. the method for predicting the cooked quality of roasted beef based on raw meat near-infrared spectrum information according to claim 1, is characterized in that, described near-infrared spectrum data model obtains by following method: b1). After establishing a large sample data of the near-infrared spectrum of raw beef for barbecue, it is tested for parameters related to the cooked quality; b2). Establish a near-infrared spectrum data model according to the corresponding relationship between the large sample data of the infrared spectrum and the parameters related to the cooked quality. 3. The method for predicting the cooked quality of roasted beef based on the near-infrared spectrum information of raw meat according to claim 2, wherein the parameter related to the cooked quality is a sensory score. 4. The method for predicting the cooked quality of roasted beef based on the near-infrared spectrum information of raw meat according to claim 3, wherein the state of the raw beef for roasting is minced meat. 5. The method for predicting the cooked quality of grilled beef based on the near-infrared spectrum information of raw meat according to claim 3, characterized in that, the raw beef for grilling is taken from the outer spine and gherkin strips. 6. The method according to any one of claims 1 to 5 for predicting the quality of roasted beef after cooking based on the near-infrared spectrum information of raw meat, characterized in that the establishment method of the near-infrared spectrum data model is the discriminant partial least squares method . 7. The method for predicting the cooked quality of grilled beef based on the near-infrared spectrum information of raw meat according to any one of claims 1 to 5, characterized in that, in the method, the obtained near-infrared spectrum is in the range of 1000-2500nm within the spectrum. 8. The method for predicting the cooked quality of grilled beef based on the near-infrared spectrum information of raw meat according to any one of claims 1 to 5, characterized in that, in the method, the method for obtaining the near-infrared spectrum is to use near-infrared The spectrometer is scanned and acquired, and the average number of scans is 30 times. 9. The method for predicting the cooked quality of grilled beef based on the near-infrared spectrum information of raw meat according to claim 8, wherein the near-infrared spectrometer is a SupNIR-1550 portable near-infrared spectrometer. 10. The application of the method for predicting the cooked quality of roasted beef based on the near-infrared spectrum information of raw meat according to any one of claims 1 to 9 in rapid grading of raw beef for roasting.