Frozen food safety detection method and detection system

Abstract

The invention provides a frozen food safety detection method and detection system, and relates to the technical field of metering detection, and the method comprises the following steps: cutting frozen food to be detected to extract a sample to be detected, and collecting safety detection item data of external measurement data points; and setting a safety detection item standard value of normal food, and obtaining the variation index of the external measurement data point in the frozen food to be detected. And acquiring the red-green axis chroma of each external measurement data point, and judging the overall corruption degree of the frozen food to be detected. And outputting the risk level to a detection report system. Spatial weight sampling and angular point / edge / center differentiated point distribution in an unfrozen state are realized through the in-situ sampling and multi-dimensional detection module, decay acceleration and sample distortion caused by traditional unfreezing detection are avoided by combining multi-index synchronous measurement, discrete detection data are quantized into a unified decay index, and the detection accuracy is improved. The problem that a traditional single threshold value judgment method is not sensitive to local corruption is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of metrological detection, in particular to a frozen food safety detection method and system. BACKGROUND

[0002] Frozen food of beef and mutton is rich in protein and water, and is easily affected by temperature fluctuations during frozen storage and transportation, leading to the reproduction of microorganisms and the decomposition of protein on the surface, especially on the exposed areas such as the packaging edges and cutting surfaces, thereby causing local spoilage. Spoilage can cause an increase in the total number of colonies and specific pathogenic bacteria, abnormal fluctuations in pH value, a significant increase in biological amine content, an increase in total volatile basic nitrogen (TVB-N), and changes in color and other parameters. The pH value can directly reflect abnormal muscle lactic acid metabolism, and total volatile basic nitrogen (TVB-N) is a sensitive indicator of protein degradation. Biological amines, especially histamine and tyramine, are toxic products of microbial activity in high-protein meat. Therefore, by detecting the above-mentioned main parameters, the degree of abnormality of the frozen food of beef and mutton can be effectively determined.

[0003] In the prior art, the disclosure number CN107271540A discloses an online monitoring method for frozen food based on the Internet. The technology includes: quantifying defects by using a formula; using the detection results of magnetic memory to evaluate the thawing status, color, and odor changes of frozen food; obtaining the maximum safety score of frozen food; using a fuzzy comprehensive evaluation system to combine quantitative and qualitative analysis, establishing an evaluation set based on engineering practice, and establishing a comprehensive evaluation judgment matrix. This is beneficial for convenient monitoring and improves the degree of intelligence. The present application is more strict in quality control of food and reduces the influence of human subjective factors. The qualified rate of products leaving the factory is higher.

[0004] However, the safety detection of frozen food of beef and mutton in the above-mentioned prior art usually relies on random sampling and thawing detection, such as taking only the central part, which causes the sample to be unable to truly reflect the spoilage condition of the high-risk area on the surface of the food, and the thawing process accelerates the reproduction of microorganisms, causing the detection data to be distorted. Most technologies only use a single indicator for judgment, lack multi-parameter collaborative analysis, and are not sensitive to early spoilage or local deterioration. The data analysis relies on artificial experience threshold, and does not establish a spatial weight and dynamic normalization algorithm, which cannot quantify the degree of spoilage and has a high misjudgment rate. It cannot locate the spoilage area or identify abnormal data. The overall process has low automation.

[0005] The above information disclosed in the background section is only used to enhance the understanding of the background of the present disclosure, and therefore it can include information that does not constitute prior art known to those of ordinary skill in the art. SUMMARY The application aims to provide a frozen food safety detection method and system to solve the problems in the background art. The application realizes spatial weight sampling in an unfrozen state through in-situ sampling and a multi-dimensional detection module, differential point distribution of corners / edges / centers, and simultaneous measurement of multiple indicators such as pH value, total volatile basic nitrogen (TVB-N), and biological amine, thereby avoiding sample distortion and corruption acceleration caused by traditional unfreezing detection, quantizing discrete detection data into a unified corruption index, and solving the problem of local corruption insensitivity of traditional single threshold determination method. The credibility verification and chroma correlation module outputs credibility Q by analyzing the difference between the red-green axis chroma difference and the abnormality index difference, effectively identifies abnormal detection points, and guides supplementary sampling, thereby overcoming the defects of high misjudgment rate and data island in conventional technologies.

[0006] To achieve the above-mentioned purpose, the application provides the following technical solutions. A frozen food safety detection method, comprising the following specific steps: S1: cutting the frozen food to be measured to extract a sample to be measured, the sample to be measured forming a cubic shape with an edge length of N centimeters, the region of the frozen food to be measured in contact with air being marked as an exposed region, and K sampling points uniformly distributed in the region exposed to air being extracted from the sample to be measured as external measurement data points, safety detection item data of the external measurement data points being collected, the safety detection items including pH value, total volatile basic nitrogen content, and biological amine content data, and representative sampling being completed; S2: setting the safety detection item standard value of normal food as the standard threshold value of pH value, total volatile basic nitrogen content, and biological amine content, establishing a safety detection mathematical model of the frozen food to be measured, calculating the absolute difference value between the safety detection item data of each external measurement data point and the standard value, and importing the safety detection mathematical model to obtain the abnormality index of the external measurement data points in the frozen food to be measured; S3: collecting the red-green axis chroma of each external measurement data point, comparing the red-green axis chroma difference and the abnormality index difference between different external measurement data points, calculating the credibility of the abnormality index, further selecting internal measurement data points in the sample to be measured, collecting the red-green axis chroma difference of the internal measurement data points, and judging the overall corruption degree of the frozen food to be measured; S4: generating a unified safety score through a preset threshold algorithm based on the maximum value of the abnormality index and the red-green axis chroma difference of the external measurement data points and the internal measurement data points, outputting a result in the form of a risk level, and adding a corruption position distribution map to guide targeted processing; and the result output is preferably integrated into a detection report system.

[0007] Further, a representative sample is extracted by cutting in step S1, and the frozen food to be measured includes beef and mutton. The cutting is performed along the direction of the largest exposed surface of the frozen food to be measured. The surface of the exposed area is divided into K equal-area regions using a systematic grid method. An intermediate point is selected in each equal-area region to obtain K external measurement data points. The area of each sampling point is a square region of Further, the environmental temperature and humidity during sampling are recorded during the safety detection process to calibrate the data. Different edge regions and corner point regions of the cubic sample to be measured are assigned sampling weights to improve the representativeness, and a sampling point position diagram is labeled.

[0008] Further, the sampling process of the safety detection project includes the following steps: after selecting the sampling point on the surface of the unfrozen cubic sample to be measured, a trace amount of sample is scraped at a fixed depth, and the scraping range is a circular region with an area of The center point of the circular region is located at the center position of each sampling point, and the scraping range meets the following conditions: Further, the safety detection mathematical model for calculating the anomaly index is as follows: Wherein: is the anomaly index of the sampling point numbered is the spatial weight of the sample to be measured on the frozen food to be measured, and the spatial weight includes the corner point region weight , the edge region weight , and the center region weight Further, after calculating the anomaly indexes of the K sampling points, the overall anomaly index mean of the sample to be measured is calculated by the following formula: Wherein: is the overall anomaly index mean of the sample to be measured; is the total number of measurement data points; is the corruption degree amplification coefficient. Further, the process of collecting the red-green axis chroma of each external measurement data point includes the following steps: after positioning the measurement data point, the chroma data of the red axis and the green axis of the measurement data point is measured continuously for 3 times, and the average value is taken as the final chroma value of the measurement data point. At the same time, the ambient light compensation value is recorded. If the numerical mutation of the red axis and the green axis between adjacent points exceeds 15%, additional encryption measurement points are added at the measurement data point. All data are labeled with three-dimensional coordinate positions and stored as a chroma distribution map with weight marks. ​

[0009] Further, in the step S3, two groups of external measurement data points are selected, and the red-green axis chroma difference between the two groups of external measurement data points is calculated, and the calculation formula is: Wherein, is the red-green axis chroma difference between the external measurement data points and the external measurement data points . is the red axis chroma value of the external measurement data point . is the red axis chroma value of the external measurement data point . is the green axis chroma value of the external measurement data point . is the green axis chroma value of the external measurement data point ; at the same time, the anomaly index difference between the external measurement data point and the external measurement data point is calculated: Wherein, is the anomaly index difference between the external measurement data point and the external measurement data point . is the anomaly index of the sampling point position number ; the correlation model between and is established by linear regression analysis, and the reliability of the overall anomaly index mean of the sample to be tested is calculated. Further, the reliability of the anomaly index of the sample to be tested is calculated by the following formula: Wherein: is the overall anomaly index reliability of the sample to be tested of the anomaly index of the sampling point position number and the sampling point position number , and then T groups of different sampling point position numbers are randomly extracted, and the average value of the anomaly index reliability corresponding to each group of sampling point position numbers is calculated, which is calculated by the following formula: Wherein, is the average value of the anomaly index reliability corresponding to the T groups of sampling point position numbers; The sampling point position is one of the sampling point positions in the T group. Further, in step S4, the anomaly index and the red-green axis chroma difference of the external measurement data points and the internal measurement data points are integrated, a unified safety score is generated by a threshold algorithm, the system automatically maps the score to a four-level risk level, and the four-level risk level includes: first-level green safety, second-level yellow attention, third-level orange warning, and fourth-level red danger. At the same time, a corruption location distribution heat map is generated based on the three-dimensional coordinate data of all measurement points, and the heat map is marked with a gradient color The value is high or low, and the corner area and the edge area are highlighted separately. The results are directly integrated into the detection report system. A frozen food safety detection system for performing the above detection method, the detection system comprises: The in-situ sampling module cuts the frozen food to be tested to extract a sample to be tested, the sample to be tested forms a cubic shape with an edge length of N centimeters, the region of the frozen food to be tested that is in contact with air is marked as an exposed region, and K uniformly distributed sampling points are extracted from the region exposed to air on the sample to be tested as external measurement data points. Collect safety detection item data of external measurement data points, the safety detection items include pH value, volatile base nitrogen content and biological amine content data, complete representative sampling; the anomaly index dynamic calculation module sets the safety detection item standard value of normal food as the standard threshold value of pH value, volatile base nitrogen content and biological amine content, establishes a safety detection mathematical model of the frozen food to be tested, calculates the absolute difference between the safety detection item data of each external measurement data point and the standard value, and imports the safety detection mathematical model to obtain the anomaly index of the external measurement data points of the frozen food to be tested; The credibility verification and chroma correlation module collects the red-green axis chroma of each external measurement data point, compares the red-green axis chroma difference between different external measurement data points and the anomaly index difference, calculates the anomaly index credibility, and further selects internal measurement data points in the sample to be tested and collects the red-green axis chroma difference of the internal measurement data points. Determine the overall degree of corruption of the frozen food to be tested; The risk decision module generates a unified safety score by a preset threshold algorithm based on the maximum value of the anomaly index and the red-green axis chroma difference of the external measurement data points and the internal measurement data points, and outputs the result as a risk level, and can be attached with a corruption location distribution map to guide targeted processing; wherein, the result output is preferably integrated into the detection report system.

[0010] Compared with the prior art, the beneficial effects of the present application are: the present application realizes spatial weight sampling in an unfrozen state through in-situ sampling and a multi-dimensional detection module, corner / edge / center differential distribution, combined with synchronous measurement of multiple indexes such as pH value, TVB-N and biological amine, to avoid the accelerated corruption and sample distortion caused by traditional unfreezing detection; the abnormality index dynamic calculation module introduces a weighted normalization algorithm to quantize the discrete detection data into a unified corruption index Y, solving the problem of local corruption insensitivity of the traditional single threshold determination method; The credibility verification and chroma correlation module outputs the credibility Q through analysis of the red-green axis chroma difference and the abnormality index difference, effectively identifies abnormal detection points and guides supplementary sampling, overcoming the defects of high misjudgment rate and data island in conventional technology; the dynamic threshold algorithm of the final risk decision and visualization report module generates a safety score and a corruption heat map, which improves the efficiency of artificial sampling inspection by more than 80%, and can accurately locate the corruption area (such as the high-risk area of the corner), providing a basis for food processing. BRIEF DESCRIPTION OF DRAWINGS Fig. 1 It is the overall flowchart of the frozen food safety detection method of the present application; Fig. 2 It is the principle block diagram of the frozen food safety detection system. DETAILED DESCRIPTION

[0011] In order to make the purpose, technical scheme and advantages of the present application clearer and more apparent, the present application will be further described in detail below with specific examples.

[0012] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present application should be understood as the usual meaning understood by those skilled in the art to which the present application belongs. The "first", "second" and similar words used in the present application do not represent any order, quantity or importance, but are only used to distinguish different components. "Include" or "contain" and similar words mean that the elements or objects before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connected" or "connected" and similar words are not limited to physical or mechanical connection, but can include electrical connection, whether direct or indirect. "Up", "down", "left", "right" and the like only represent relative positional relationship, when the absolute position of the described object changes, the relative positional relationship may also change accordingly. EMBODIMENT

[0013] Please refer to Figs. 1-2 The present application provides the following technical solutions: A frozen food safety detection method, the specific steps include: S1: A representative sample is extracted by precisely cutting the frozen food to be tested. The sample is shaped into a standard cube with a fixed edge length of N centimeters. In this embodiment, N is 20 centimeters to ensure regularity and consistency. The surface layer area in direct contact with the air is particularly preserved and marked as the exposed area to accurately reflect the part of the frozen beef and mutton food exposed to the external environment. Then, K sampling points are extracted on the exposed area using a uniform distribution strategy. These points serve as external measurement data points for subsequent data collection. The safety detection item data collected from these external measurement data points include pH value (for assessing acid-base changes), volatile nitrogen content (for indicating protein spoilage), and biological amine content data (for monitoring microbial metabolic activity). Through the above process, a representative sampling link is completed, ensuring that the sample truly reflects the overall state of the frozen food. During the entire sampling process, the sample remains in an unfrozen state and is measured in situ at its original position to avoid introducing detection errors due to the thawing process. Precise cutting is performed on the frozen food to be tested to extract a highly representative sample. The frozen food to be tested mainly includes beef and mutton products. During cutting, the maximum exposed surface direction of the frozen food to be tested is strictly followed to ensure the integrity and maximization of the exposed area. Subsequently, a systematic grid division method is used to divide the entire surface into K areas of equal area. In this embodiment, K is 25, so each sampling area is 16 cm². In each equal-area region, the geometric center point is carefully selected as the sampling point, and K external measurement data points are accurately obtained. Finally, the coverage range of each sampling point is defined as a square area with an area of During the collection of safety detection item data, the environmental temperature and humidity details at the time of sampling are recorded in detail. These parameters are used for subsequent data calibration to ensure the accuracy and reliability of the detection results. Meanwhile, specific sampling weight values are assigned to the cube shape of the sample based on the special characteristics of different edge regions and corner regions to effectively improve the representativeness and detection accuracy of the sample. In this process, the specific location diagram of the sampling points is clearly marked to visually display the point distribution, supporting subsequent analysis and evaluation. The sampling process of the safety detection item includes: after selecting the sampling points on the surface of the unfrozen cube sample, a small amount of sample is scraped from the surface layer at a fixed depth, and the scraping range is a circular area with an area equal to the center point of the circular area is located at the center position of each sampling point, and the scraping range Need to meet: S2: set the normal food safety detection project standard value as the standard threshold value of pH value, volatile base nitrogen content and biological amine content, establish the safety detection mathematical model of the frozen food to be measured, calculate the absolute difference value of the safety detection project data in each external measurement data point and the standard value, and import the safety detection mathematical model to obtain the anomaly index of the external measurement data point in the frozen food to be measured; The safety detection mathematical model for calculating the anomaly index is as follows: Wherein: is the anomaly index of the sampling point number , i=1, 2,..., K; a, b, g , respectively, are the standard values of pH value, volatile base nitrogen and biological amine; a, b, g , respectively, are the absolute difference values between the measured values and the standard values of pH value, volatile base nitrogen and biological amine; , respectively, are the project weight coefficients of pH value, volatile base nitrogen and biological amine Meet the following relationship: And also need to meet: α+β+γ=1; is the spatial weight of the sample to be measured on the frozen food to be measured, and the spatial weight Includes: the corner point area weight of the frozen food , the edge area weight and the center area weight , and the three spatial weights meet the following relationship: After calculating the anomaly index of the K sampling points, the overall anomaly index average of the sample to be measured is calculated by the following formula: Wherein: is the overall anomaly index average of the sample to be measured, The value of the result is determined by the anomaly index value calculated from each measurement data point; is the total number of measurement data points, The overall anomaly index average is inversely proportional; is the corruption degree amplification coefficient, The overall anomaly index average is proportional, and the value of in the embodiment is 100. The process of collecting the red-green axis chroma of each external measurement data point specifically includes the following steps: First, accurately position the spatial position of each external measurement data point to ensure accurate measurement points; Subsequently, for each located data point, three independent chroma data acquisitions are performed in succession, and the specific values of red-axis chroma and green-axis chroma are measured respectively to obtain stable and reliable original data; after the measurement is completed, the arithmetic mean values of the three groups of red-axis chroma and green-axis chroma values are calculated respectively, which are used as the final red-axis chroma value and the final green-axis chroma value of the measurement data point, so as to improve the representativeness and accuracy of the data; At the same time, during the entire measurement process, the compensation value of the ambient light must be recorded in real time for correcting the interference of the chroma data caused by the change of external light conditions; In addition, during the data processing stage, if the value change rate (i.e. the mutation amplitude) of the red-axis chroma value or the green-axis chroma value between adjacent measurement points exceeds 15%, three additional encryption measurement points are immediately supplemented at the measurement point position for repeated measurement to verify the reliability of the data; All collected chroma values and related information need to be marked with their accurate three-dimensional coordinate positions to ensure that each data point has a clear spatial positioning; Finally, all the above data (including chroma values, coordinate positions and weight marks) are integrated and stored as a structured chroma distribution map with weight marks for subsequent data analysis and visual presentation. S3: Collect the red-green axis chroma of each external measurement data point, compare the red-green axis chroma difference and the heterogeneity index difference between different external measurement data points, calculate the heterogeneity index reliability, and further select internal measurement data points in the sample to be measured, and collect the red-green axis chroma difference of the internal measurement data points to judge the overall spoilage degree of the frozen food to be measured; Specifically, after collecting the red-axis chroma and green-axis chroma data of each external measurement data point, the color difference between multiple external measurement data points and the spoilage characteristics need to be analyzed. The specific steps include: First, two different groups of external measurement data points are systematically selected, and the red-axis chroma difference and the green-axis chroma difference between them are calculated to obtain a comprehensive red-green axis chroma difference quantitative index; At the same time, the difference value of the heterogeneity index between the same two groups of measurement points is accurately compared, which directly reflects the discrete characteristics of the spoilage degree at different positions; Subsequently, the two key data, i.e. the red-green axis chroma difference and the heterogeneity index difference, are statistically correlated and modeled through linear regression analysis to establish a correlation model between the two; Based on the model, the overall reliability evaluation value of the heterogeneity index data set of all measurement points is calculated to judge the reliability of the sampling data; After the external data verification is completed, further select a plurality of internal measurement data points with spatial representation in the non-exposed area of the sample to be tested, i.e. inside the cube, and collect the red axis chroma and green axis chroma data of these internal points using the same technical specification; Based on the internal point data, the red-green axis chroma difference is calculated again, and compared with the spatial correlation of the external data results; Finally, through the coupling analysis of internal and external data, combined with the reliability verification results, the scientific judgment of the spoilage degree distribution characteristics of the frozen food to be tested in the overall structure is obtained Select two groups of external measurement data points, and calculate the red-green axis chroma difference between the two groups of external measurement data points, the calculation formula is: Wherein, is the red-green axis chroma difference between the external measurement data points and the external measurement data points , which is used to reflect the degree of color mutation; is the red axis chroma value of the external measurement data point ; is the red axis chroma value of the external measurement data point ; is the green axis chroma value of the external measurement data point ; is the green axis chroma value of the external measurement data point , each chroma value is collected by a colorimeter; at the same time, the heterogeneity index difference between the external measurement data points and the external measurement data points is calculated: Wherein, is the heterogeneity index difference between the external measurement data points and the external measurement data points ; is the heterogeneity index of the sampling point with number ; the correlation model between and is established by linear regression analysis, and the reliability of the overall heterogeneity index mean of the sample to be tested is calculated. The reliability of the heterogeneity index of the sample to be tested is calculated by the following formula: Wherein: is the heterogeneity index of the sampling point with number and the sampling point with number ; then randomly select T groups of different sampling point numbers, calculate the average value of the heterogeneity index reliability corresponding to each group of sampling point numbers , and calculate it by the following formula: Wherein, is the average value of the heterogeneity index reliability corresponding to the T groups of sampling point numbers; is a variable is a structure of taking values from 1 to T in sequence according to P and summing up. is a sampling point in one of the T groups of sampling points, as T increases, the result gradually decreases. S4: the maximum value of the abnormality index in the external measurement data points and the internal measurement data points and the red-green axis chroma difference value are generated into a unified safety score through a preset threshold algorithm, and the output result is a risk level, and a corruption location distribution map can be added to guide targeted processing; wherein, the result output is preferably integrated into a detection report system.

[0014] All key indicators collected by the external measurement data points and the internal measurement data points are comprehensively integrated, including the abnormality index of each point, the quantified value reflecting the degree of corruption, and the red-green axis chroma difference value representing the difference value of color change; then, the integrated data are calculated and processed through a preset threshold algorithm to generate a unified safety score value, and the algorithm automatically judges the corruption risk level based on a preset rule; After obtaining the safety score, the system automatically maps it to a preset four-level risk level classification system: First-level green safety: indicating that the food is in an ideal safe state; Second-level yellow attention: indicating that there is a slight risk that needs to be monitored; Third-level orange warning: prompting that the medium corruption needs to be handled in time; Fourth-level red danger: warning that the serious corruption must be intervened immediately; At the same time, using the three-dimensional coordinate data recorded by all measurement points, including spatial position information, the system dynamically generates a corruption location distribution heat map, which uses a gradient color scheme to intuitively mark the high and low degrees of the abnormality index of each point, and the corner areas and edge areas are highlighted separately to highlight these high-risk areas; Finally, all the above analysis results, including the safety score, the risk level and the heat map, are directly integrated into the detection report system to realize automatic output and visual presentation, so that the user can quickly obtain and apply the detection conclusion. The embodiment also provides a frozen food safety detection system, which is used to execute the detection method described above, and the detection system comprises: An in-situ sampling module cuts the to-be-tested frozen food to extract a to-be-tested sample, the to-be-tested sample forms a cubic shape with an edge length of N centimeters, a region of the to-be-tested frozen food in contact with air in the to-be-tested sample is marked as an exposed region, and K uniformly distributed sampling points in the region exposed to air from the to-be-tested sample are extracted as external measurement data points, safety detection item data of the external measurement data points are collected, the safety detection items include pH value, volatile base nitrogen content and biological amine content data, and representative sampling is completed. The abnormality index is dynamically calculated by setting the safety detection item standard value of normal food as the standard threshold value of pH value, volatile base nitrogen content and biological amine content, establishing a safety detection mathematical model of the frozen food to be measured, calculating the absolute difference between the safety detection item data and the standard value in each external measurement data point, and importing the safety detection mathematical model to obtain the abnormality index of the external measurement data point of the frozen food to be measured; The reliability verification and chroma correlation module collects the red-green axis chroma of each external measurement data point, compares the red-green axis chroma difference and the abnormality index difference between different external measurement data points, calculates the abnormality index reliability, further selects an internal measurement data point in the sample to be measured, collects the red-green axis chroma difference of the internal measurement data point, and judges the overall spoilage degree of the frozen food to be measured; the risk decision module generates a unified safety score through a preset threshold algorithm by using the maximum value of the abnormality index and the red-green axis chroma difference of the external measurement data point and the internal measurement data point, outputs a result as a risk level, and can additionally attach a spoilage position distribution map to guide targeted processing; wherein, the result output is integrated into a detection report system.

[0015] The above formulas are all dimensionless numerical calculations, the formulas are obtained by collecting a large amount of data to simulate a formula of the nearest real situation, and the preset parameters in the formula are set by a person skilled in the art according to the actual situation.

[0016] The above embodiments can be realized by software, hardware, firmware or any combination thereof, in whole or in part. When realized by software, the above embodiments can be realized in the form of a computer program product in whole or in part. Those skilled in the art can realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized by hardware or software methods depends on the specific application and design constraints of the technical solutions. The units described as separate components can be or can not be physically separated, the components displayed as units can be or can not be physical units, and can be located in one place or distributed on multiple network units. Part or all of the units can be selected according to actual needs to realize the purpose of the embodiments.

[0017] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered within the protection scope of the present application.

Claims

1. A method for frozen food safety detection, characterized in that, The specific steps include: S1: cutting the frozen food to be tested to extract a sample to be tested, the sample to be tested forming a cubic shape with an edge length of N centimeters, the region of the frozen food to be tested in contact with air being marked as an exposed region, and K sampling points being extracted from the region exposed to air on the sample to be tested as external measurement data points, safety detection item data of the external measurement data points being collected, the safety detection items including pH value, volatile base nitrogen content, and biological amine content data, and representative sampling being completed; S2: setting a normal food safety detection item standard value as a standard threshold value of pH value, volatile base nitrogen content, and biological amine content, establishing a safety detection mathematical model of the frozen food to be tested, calculating the absolute difference between the safety detection item data of each external measurement data point and the standard value, and importing the safety detection mathematical model to obtain the anomaly index of the external measurement data points of the frozen food to be tested; S3: collecting the red-green axis chroma of each external measurement data point, comparing the red-green axis chroma difference and the anomaly index difference between different external measurement data points, calculating the anomaly index reliability, further selecting internal measurement data points in the sample to be tested, collecting the red-green axis chroma difference of the internal measurement data points, and judging the overall spoilage degree of the frozen food to be tested; S4: generating a unified safety score through a preset threshold algorithm based on the maximum value of the anomaly index and the red-green axis chroma difference of the external measurement data points and the internal measurement data points, outputting a result as a risk level, and adding a spoilage position distribution map to guide targeted processing; wherein the result output is preferably integrated into a detection report system.

2. The method of claim 1, wherein: A representative sample is extracted by cutting the frozen food to be measured in step S1, which includes beef and mutton, and the cutting is performed along the direction of the largest exposed surface of the frozen food to be measured. On the surface of the exposed area, the surface is divided into K equal-area regions using a systematic grid method, and a middle point is selected in each equal-area region to obtain K external measurement data points, and the area of each sampling point is a square region of ​ During the collection of safety detection items, the environmental temperature and humidity at the time of sampling are recorded to calibrate the data; the sampling weights of different edge regions and corner regions of the cubic sample to be tested are allocated to improve the representativeness, and a sampling point position diagram is marked.

3. The method of claim 2, wherein the method is a method of frozen food safety detection. The sampling process of the safety detection item includes: after selecting sampling points on the surface of the unfrozen cubic sample to be tested, scraping a trace sample of a fixed depth of the surface layer, and the scraping range is a circular area with an area equal to The center point of the circular area is located at the center position of each sampling point, and the scraping range needs to meet: .

4. The method of claim 2, wherein the method comprises: The security detection mathematical model for calculating the mutation index is as follows: Wherein: is the heterogeneity index of the sampling point with the point number ; are the standard values of pH, volatile basic nitrogen and biogenic amine, respectively; are the absolute differences between the measured values and the standard values of pH, volatile basic nitrogen and biogenic amine, respectively; are the item weight coefficients of pH, volatile basic nitrogen and biogenic amine, respectively, satisfy the following relationship: ; is the spatial weight of the sample to be measured on the frozen food to be measured, and the spatial weight includes: the corner point area weight , the edge area weight and the center area weight of the frozen food to be measured.

5. The method of claim 4, wherein: After calculating the heterogeneity index of K sampling points, the average heterogeneity index of the whole sample is calculated by the following formula: Wherein: The overall heterogeneity index mean for the sample under test; to measure the total number of data points; is the degree of corruption amplification factor.

6. The method of claim 4, wherein the method comprises, The process of collecting the red-green axis chroma of each external measurement data point includes: after positioning the measurement data point, continuously measuring the chroma data of the red axis and the green axis of the measurement data point for 3 times, taking the average value as the final chroma value of the measurement data point, recording the ambient light compensation value, if the numerical mutation of the red axis and the green axis between adjacent points exceeds 15%, supplementing multiple encrypted measurement points at the measurement data point, synchronously marking the three-dimensional coordinate positions of all data, and storing them as a color distribution map with weight marks.

7. The method of claim 6, wherein the method comprises: In the step S3, two groups of external measurement data points are selected, and the red-green axis chroma difference value between the two groups of external measurement data points is calculated, and the calculation formula is: Wherein, is the red-green axis chroma difference value between the external measurement data point and the external measurement data point ; is the red axis chroma value of the external measurement data point ; external measurement data points red axis chroma value; external measurement data points green axis chroma value; the green axis chroma value of the external measurement data point ; meanwhile, the difference in the heterogeneity index between the external measurement data point and the external measurement data point is calculated: wherein, the difference in the heterogeneity index between the external measurement data point and the external measurement data point ; the heterogeneity index of the sampling point with the point number ; the correlation model between and is established through linear regression analysis, and the reliability of the overall heterogeneity index mean value of the sample to be tested is calculated.

8. The method of claim 7, wherein the method comprises, The credibility of the heterogeneity index of the sample to be tested is calculated by the following formula: Wherein: is the heterogeneity index of the sample to be tested at the sampling point position numbered and the sampling point position numbered The credibility of the heterogeneity index of the sample to be tested is calculated by the following formula: Wherein: is the average value of the credibility of the heterogeneity index corresponding to each group of sampling point position numbers T groups of different sampling point position numbers are randomly selected, and the average value of the credibility of the heterogeneity index corresponding to each group of sampling point position numbers is calculated is the sampling point position of one of the T groups of sampling point positions.

9. The method of claim 1, wherein: In step S4, the anomaly index and the red-green axis chroma difference of the integrated external measurement data points and internal measurement data points are generated by a threshold algorithm to generate a unified safety score, and the system automatically maps the score to a four-level risk level, including: first-level green safety, second-level yellow attention, third-level orange warning, and fourth-level red danger. At the same time, a corruption location distribution heat map is generated based on the three-dimensional coordinate data of all measurement points, and the heat map is marked with a gradient color The value is high or low, and the corner area and the edge area are highlighted separately. The result is directly integrated into the detection report system.

10. A frozen food safety inspection system characterized by comprising: The detection system is used to perform the detection method of any one of claims 1-9, and the detection system comprises: an in-situ sampling module, which cuts the frozen food to be tested to extract a sample to be tested, the sample to be tested forming a cubic shape with an edge length of N centimeters, the region of the frozen food to be tested in contact with air being marked as an exposed region, and K sampling points being extracted from the region exposed to air on the sample to be tested as external measurement data points, safety detection item data of the external measurement data points being collected, the safety detection items including pH value, volatile base nitrogen content, and biological amine content data, and representative sampling being completed; The abnormality index is dynamically calculated by setting the standard value of the safety detection item of normal food as the standard threshold value of pH value, volatile base nitrogen content and biological amine content, establishing a safety detection mathematical model of the frozen food to be measured, calculating the absolute difference between the safety detection item data and the standard value in each external measurement data point, and importing the safety detection mathematical model to obtain the abnormality index of the external measurement data point in the frozen food to be measured; The reliability verification and chroma correlation module collects the red-green axis chroma of each external measurement data point, compares the red-green axis chroma difference and the abnormality index difference between different external measurement data points, calculates the abnormality index reliability, further selects the internal measurement data point in the sample to be measured, collects the red-green axis chroma difference of the internal measurement data point, and judges the overall spoilage degree of the frozen food to be measured; The risk decision module generates a unified safety score through a preset threshold algorithm by using the maximum value of the abnormality index and the red-green axis chroma difference in the external measurement data point and the internal measurement data point, outputs the result as a risk level, and can additionally attach a spoilage position distribution map to guide targeted processing; wherein the result output is preferably integrated into a detection report system.

Images