Method for evaluating beef freshness based on near infrared spectrum technology

Abstract

The invention discloses a method for evaluating beef freshness based on a near infrared technology, and belongs to the technical field of fresh meat detection. The method for evaluating the beef freshness comprises the following steps: establishing a quantitative prediction model by utilizing near infrared spectrum information and physicochemical properties and quality indexes of fresh beef, so as to quickly evaluate the beef freshness based on a near infrared spectrum technology; the near infrared spectrum information is the spectrum information of the surface of the fresh beef within the range of 1000-1800 nm; the physicochemical properties comprise the pH value and the TVB-N content; the quality indexes comprise chromatic aberration L *, a * and b * values. According to the method, the beef freshness can be rapidly detected, data tests prove that the accuracy is high, the practicability is good, the limitation of human eye recognition is overcome, and rapid data freshness identification is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of fresh meat detection, and particularly relates to a method for evaluating the freshness of beef based on near-infrared spectroscopy technology. Background Art

[0002] Beef is a food matrix with high nutritional value, rich in various essential amino acids, fatty acids, vitamins and minerals. Its unique flavor characteristics have earned it a high reputation among consumers. With the gradual improvement of the living standards of residents, the market consumption of beef shows a significant growth trend, especially the demand for high-quality and high-nutritional-value beef is increasing day by day. The freshness of meat, as a core parameter for evaluating the quality of beef, is crucial for ensuring food safety. During the storage of beef, the microbial activity increases, leading to a large number of reproductions and the accompanying generation of various metabolic end products. This process promotes the spoilage of the meat quality, thereby affecting the sensory quality of beef and the health and safety of consumers. Therefore, the food safety issue of beef during storage has become the focus of attention in the academic and industrial circles.

[0003] In view of the complex process, low efficiency, and large consumption of time and human resources of traditional beef freshness detection methods, these methods are difficult to meet the needs of real-time monitoring of meat freshness in daily production and life. Therefore, there is an urgent need to develop a rapid, simple, scientific and objective detection method.

[0004] Near-infrared spectroscopy technology, as a rapid, non-destructive, real-time monitoring and environmentally friendly analysis method, has been widely used in the field of rapid assessment of food freshness and safety. Although near-infrared spectroscopy technology has been widely adopted in the field of food quality monitoring, especially in freshness assessment, however, when this technology relies on a single index such as total volatile basic nitrogen (TVB-N) to determine the freshness of beef, the effectiveness of its application is still limited, and the accuracy of the evaluation results is still insufficient. Summary of the Invention

[0005] Aiming at the defects existing in the prior art, the purpose of the present invention is to provide a method for evaluating the freshness of beef based on comprehensive evaluation of multiple indicators. This method comprehensively measures the characteristics of the prediction model by parameters such as TVB-N, PH and chromaticity, combines the laws of model data, and adopts similar preprocessing forms and sample set division modes to construct an evaluation model for beef freshness to achieve accurate evaluation of beef freshness.

[0006] In order to achieve the above purpose, the following technical solutions are provided:

[0007] The present invention provides a method for evaluating the freshness of beef based on near-infrared spectroscopy. The method includes establishing a quantitative prediction model using near-infrared spectral information and the physicochemical properties and quality indicators of fresh beef, so as to rapidly evaluate the freshness of beef based on near-infrared spectroscopy.

[0008] In one embodiment, the near-infrared spectral information is the spectral information of the surface of fresh beef in the range of 1000 nm to 1800 nm.

[0009] In one embodiment, the physicochemical properties include pH value and TVB-N content; the quality indicators include color difference L*, a*, b* values.

[0010] Preferably, when measuring the quality indicators of fresh beef, the state of the fresh beef sample is a meat block; when measuring the physicochemical properties of fresh beef, the state of the fresh beef sample is minced meat.

[0011] Preferably, when the physicochemical property of the beef is TVB-N, the near-infrared spectrum obtained is 1000 nm to 1800 nm;

[0012] When the physicochemical property of the beef is pH value, the near-infrared spectrum obtained is 1000 nm to 1800 nm;

[0013] When the quality indicator of the beef is chromaticity, the near-infrared spectrum obtained is 1000 nm to 1800 nm.

[0014] In one embodiment, when measuring the TVB-N content of fresh beef, the freshness grade of the meat is classified according to the detection result of the TVB-N value: if the TVB-N content is less than 15 mg / 100 g, it is determined as fresh meat; if the TVB-N content is between 15 mg / 100 g and 25 mg / 100 g, it is classified as relatively fresh meat; if the TVB-N content exceeds 25 mg / 100 g, it is identified as spoiled meat.

[0015] In one embodiment, the parameters obtained for the near-infrared spectrum are: a SupNIR-2700 type near-infrared spectrometer, with an average number of scans of 30 times, a resolution of 11 nm, and an analysis time of 30 s.

[0016] In one embodiment, the fresh beef is specifically the beef eye muscle part.

[0017] In one embodiment, the method for establishing the quantitative prediction model of the near-infrared spectrum is discriminant partial least squares method.

[0018] Furthermore, during the establishment of the quantitative prediction model, the near-infrared spectral information is processed by first derivative.

[0019] Preferably, in the process of establishing the quantitative prediction model:

[0020] When the beef physical and chemical property is TVB-N, the pretreatment method is SG;

[0021] When the beef physical and chemical property is pH, the pretreatment method is 1DSG;

[0022] When the beef quality index is L*, the pretreatment method is RAW;

[0023] When the beef quality index is a*, the pretreatment method is SG;

[0024] When the beef quality index is b*, the pretreatment method is MSC.

[0025] The present invention also provides an application of the above method in beef storage monitoring.

[0026] Beneficial effects:

[0027] The present invention adopts the fast, non-destructive and reliable NIR technology, designs a new efficient and reliable beef quality identification method, and realizes multi-faceted monitoring of beef freshness, storage time and physical and chemical indexes based on near-infrared spectral information. Its simple operation process and good repeatability not only broaden the application scope of meat freshness detection technology, but also effectively help producers and enterprises reduce resource investment in quality control, prevent economic losses caused by inferior products, and at the same time provide economic, efficient, scientific and rigorous support for beef quality evaluation and pricing strategies. Description of the drawings

[0028] Figure 1 It is the near-infrared average spectrogram of the beef sample after the first derivative treatment in the embodiment of the present invention;

[0029] Figure 2 It is the chromaticity value change data diagram of the beef eye muscle stored at 4°C for 0 to 7 days;

[0030] Figure 3 It is the TVB-N value change data diagram of the beef eye muscle stored at 4°C for 0 to 7 days;

[0031] Figure 4 It is the pH value change data diagram of the beef eye muscle stored at 4°C for 0 to 7 days. Specific embodiments

[0032] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention. The following specific implementation manners further describe the present invention.

[0033] Embodiment 1

[0034] A method for evaluating the freshness of beef based on near-infrared spectroscopy technology, comprising the following steps:

[0035] (1) Sample preparation

[0036] Place the chilled beef without any freezing in a foam insulation box equipped with ice packs and transport it to the laboratory within 1 hour; cut it into suitable-sized block samples, then place them on a black tray respectively, seal them with PE plastic wrap, and store them in a 4°C refrigerator. Select the same time point every day to take samples for spectral determination and physical and chemical index determination. Each sample is measured in parallel at least three times, and continuously measured for 7 days; such as Figures 2 to 4 ;

[0037] (2) Determination of physical and chemical indexes of beef

[0038] pH value determination: Refer to GB5009.237-2016, accurately weigh 4 g of the sample, add 40 mL of deionized water, and use a vortex oscillator to fully homogenize it; turn on the pH meter in advance, preheat it for 30 min, then insert the electrode into the sample solution and read the value; repeat three parallel experiments and take the average value as the pH value of the sample;

[0039] TVB-N content determination: The specific method refers to the method of GB5009.228-2016 "Determination of Volatile Basic Nitrogen in Foods", using the automatic Kjeldahl method. Accurately weigh 10 g of the test sample (accurate to 0.001 g) into the distillation tube, add 75 mL of deionized water and shake it to achieve uniform dispersion of the sample in the sample solution, and keep it immersed for 30 min; add 1 g of magnesium oxide, and then put it into the automatic Kjeldahl apparatus; to ensure accuracy, each sample is measured at least 3 times, and the average value is taken as the TVB-N content of the sample;

[0040] Chromaticity value determination: Use a DS-700D handheld colorimeter to measure the color of beef. In order to obtain more accurate data, ensure that the color of each beef sample is measured on 6 different surfaces of the meat block respectively, and use it as a reference value for quantitative analysis. To ensure accuracy, each sample is measured at least 6 times, and the average value is taken as the L* and a* values of the sample;

[0041] (3) Obtain spectral data

[0042] After removing the visible connective tissue, fat, and fascia from the beef samples, use a meat grinder to crush the samples (stir for 30 s); use a SupNIR-2700 near-infrared spectrometer to collect the near-infrared spectra of the beef samples in diffuse reflection mode; when starting to scan the samples, it is necessary to preheat the near-infrared spectrometer in advance for about 30 min, test the instrument performance, and after the reference operation, evenly spread the minced meat on the test dish, and use a SupNIR-2700 near-infrared spectrometer for measurement. Collect the spectra under a dark background using a grating dispersion type portable near-infrared spectrometer, with a wavelength range of 1000 - 1800 nm, a resolution of 11 nm, an average number of scans of the instrument of 30 times, and an analysis time of 30 s. Import the collected near-infrared spectral images into data processing software to obtain spectral data; each sample is filled and measured 3 times repeatedly and the average value is taken.

[0043] (4) Pretreatment of near-infrared spectral data

[0044] The prediction accuracy of near-infrared spectra is often affected by some factors unrelated to the sample properties, such as environmental temperature, sample state, light scattering, and instrument response. These factors lead to baseline drift and poor repeatability of near-infrared spectra.

[0045] Adopt the Savitzky-Golay convolution smoothing method to smooth the data, standard normal variate transformation (SNV) to standardize the data distribution, multiplicative scatter correction (MSC) to reduce the influence of scattering effects, and derivative spectra method and other multiple pretreatment methods to highlight the subtle changes in the spectra, eliminate various high-frequency noises and baseline drift, and improve repeatability and signal-to-noise ratio.

[0046] Figure 1 Figure 14 is the near-infrared spectral diagram after first-order derivative processing. It can be seen from the figure that after the spectra are processed by first-order derivative, the linear baseline drift is effectively reduced and the spectral band characteristics are strengthened. The 0-H absorption peak at 1150 nm, the C-H absorption peak at 1300 - 1500 nm, and the N-H absorption peak at 1400 - 1600 nm in the collected spectral data are all more obvious. Among them, the near-infrared spectral absorption peaks collected under intact meat blocks fluctuate more in the range of 1000 - 1150 nm and 1500 nm. The results of the pretreatment are shown in Table 1:

[0047] Table 1. Quantitative PLS models of each index in samples established by different spectral pretreatment methods

[0048]

[0049] The results were analyzed in combination with the data in the table. MSC eliminated the scattering induced by uneven particle diameters and the influence of the linear shift in the spectrum; MeanCenter and MedianCenter removed the average value or median value from each numerical sequence to increase the deviation between sample spectra, thereby improving the measurement ability of the model and enhancing reliability; 1stD belongs to the derivative denoising mode, which can effectively eliminate the interference of the baseline and various background noises, separate overlapping peaks, and improve resolution and accuracy.

[0050] (5) Random diversity of samples

[0051] Given that the variation degrees of different samples are significantly different, the data distribution shows a normal distribution with good dispersion, all of which indicate that this sample set has a considerable degree of representativeness statistically. To verify whether near-infrared spectroscopy technology has general predictability in evaluating beef freshness indicators, a comprehensive modeling strategy was implemented. According to a ratio of 3:1, the samples in each group were divided into a calibration set and a prediction set respectively to comprehensively evaluate the accuracy and reliability of spectral analysis in predicting beef freshness indicators.

[0052] (6) Establishment of the prediction model

[0053] Selecting the beef eye as a sample, using a SupNIR-2700 near-infrared spectrometer, the near-infrared spectra of beef samples were collected in the range of 1000 - 1800 nm in diffuse reflection mode, and the main physicochemical indicators (TVB-N, pH) and quality indicators (L*, a*, b*) were measured. A partial least squares (PLS) method was used to establish a quality prediction model. Finally, good prediction effects on TVB-N and pH were obtained, with R values of 0.91 and 0.87 respectively, and good prediction effects on L*, a*, and b* were obtained, with R values of 0.82, 0.64, and 0.84 respectively.

[0054] 8. Freshness prediction

[0055] According to the above methods of sample preparation, obtaining spectral data, data preprocessing, and obtaining near-infrared spectral characteristic values, the near-infrared spectral characteristic values of the beef samples to be measured were collected, and the near-infrared spectral characteristic values were input into the established PLSR prediction model to obtain the freshness of the beef to be measured.

[0056] Although the present invention has been disclosed above with preferred embodiments, it is not intended to limit the present invention. Any person familiar with this technology can make various modifications and decorations without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention should be defined by the claims.

Claims

1. A method for evaluating the freshness of beef based on near-infrared spectroscopy technology, characterized in that, The method includes establishing a quantitative prediction model by using near-infrared spectral information and the physicochemical properties and quality indexes of fresh beef, so as to rapidly evaluate the freshness of beef based on near-infrared spectroscopy; the near-infrared spectral information is the spectral information on the surface of fresh beef in the range of 1000 nm to 1800 nm; the physicochemical properties include pH value and TVB-N content; the quality indexes include color difference L*, a*, b* values.

2. The method according to claim 1, wherein When measuring the quality indexes of fresh beef, the state of the fresh beef sample is a meat block; when measuring the physicochemical properties of fresh beef, the state of the fresh beef sample is minced meat.

3. The method according to claim 1, characterized in that, When the physicochemical properties of the beef are TVB-N and pH value, the obtained near-infrared spectrum is 1000 nm to 1800 nm.

4. The method according to claim 1, wherein When the quality index of the beef is chromaticity, the obtained near-infrared spectrum is 1000 nm to 1800 nm.

5. The method according to claim 1, wherein The parameters obtained by the near-infrared spectrum are: SupNIR-2700 type near-infrared spectrometer, the average number of scans is 30 times, the resolution is 11 nm, and the analysis time is 30 s.

6. The method according to claim 1, characterized in that, The fresh beef is specifically the beef eye muscle part.

7. The method according to claim 1, wherein The method for establishing the quantitative prediction model is the discriminant partial least squares method.

8. The method according to claim 1, characterized in that, In the process of establishing the quantitative prediction model, the near-infrared spectral information is processed by first derivative.

9. The method according to claim 1, characterized in that In the process of establishing the quantitative prediction model: When the physicochemical property of the beef is TVB-N, the pretreatment method is SG; When the physicochemical property of the beef is pH, the pretreatment method is 1DSG; When the quality index of the beef is L*, the pretreatment method is RAW; When the quality index of the beef is a*, the pretreatment method is SG; When the quality index of the beef is b*, the pretreatment method is MSC.

10. Application of the method according to any one of claims 1 to 9 in beef storage monitoring.

Images