Evaluation method and evaluation system

Abstract

To provide an analysis method and an analysis system for analyzing a food state of a fermented food, for example, without relying on sensory evaluation and simple fluorescence measurement.SOLUTION: An analysis method of the present invention comprises irradiating a food in a fermentation process, for example, with light dispersed in a visible light region from ultraviolet ray having a continuous wavelength, measuring a spectrum of fluorescence generated by the irradiation to acquire a 3D fluorescence spectrum, and comparing the spectrum with a database previously measured, wherein similarity is evaluated using multivariate analysis and artificial intelligence during comparison to determine a food state of the food.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an evaluation method and an evaluation system for evaluating characteristics of an object based on its fluorescence spectrum. [Background technology]

[0002] In recent years, three-dimensional fluorescence analysis (EEM: Excitation-Emission Matrix) has become known as a method for analyzing foods and other foods. For example, Non-Patent Document 1 discloses a method for estimating the aging and constituents of sake by utilizing differential spectra obtained by EEM. Non-Patent Document 2 also discloses a method for determining the quality characteristics of awamori, measuring the EEM of awamori and standard reagents such as vanillin and 4-vinyl guai alcohol, attributing the 3D spectra, and examining the relationship with storage period. It also discloses a method for separating new sake from aged sake (aged by storage period) using principal component analysis and score plotting. Furthermore, as shown in Patent Document 1, a known component analyzer of this type includes a data acquisition unit that acquires spectral data obtained by measuring the spectrum of an object containing multiple components with a sensor, and an analysis circuit that identifies multiple components by performing factor analysis on the input data, which are the spectral data acquired by the data acquisition unit and at least one reference spectral dataset read from a storage unit that pre-stores at least one reference spectral dataset representing the spectra of multiple substances. [Prior art documents] [Patent documents]

[0003] [Non-Patent Document 1] Aichi Industrial Science and Technology Center Food Industry Technology Center News January 2023 [Non-patent document 2] Journal of the Japanese Society for Food Science and Technology, Vol. 64, No. 12, December 2017, pp. 577-583 [Patent Document 1] Japanese Patent Publication No. 2020-34545 Summary of the Invention [Problem to be solved by the invention]

[0004] However, such conventional component analysis devices are only capable of component analysis when reference spectral data associated with feature factors can be applied, and are insufficient for obtaining general characteristics of the target object, or even complex information such as the fermentation state or taste of food and drink.

[0005] In view of the above circumstances, an object of the present invention is to provide an evaluation method and evaluation system that can evaluate general and complex information about an object. [Means for solving the problem]

[0006] The evaluation method of the first invention is a method for evaluating characteristics of an object based on a fluorescence spectrum, comprising: a data acquisition step of acquiring three-dimensional spectral data obtained by measuring the three-dimensional fluorescence spectrum of the object; a mapping step of mapping the results obtained by multivariate analysis of the three-dimensional spectral data onto a two-dimensional space; The present invention is characterized by comprising:

[0007] According to the evaluation method of the first aspect of the invention, the three-dimensional spectral data acquired in the data acquisition step is subjected to multivariate analysis. Then, by mapping the results of the multivariate analysis into two-dimensional space, the characteristics of the object, which exhibit a complex shape in the three-dimensional spectrum, can be displayed as a map in two-dimensional space as evaluation indices with clear feature quantities as the results of the multivariate analysis.

[0008] In this way, the evaluation method of the first invention can provide an index for evaluating general and complex information about an object.

[0009] The evaluation method of the second invention is the evaluation method of the first invention, The multivariate analysis is characterized by being a multidimensional scaling method.

[0010] According to the evaluation method of the second invention, by selecting multidimensional scaling as the multivariate analysis method, it is possible to easily and reliably convert the relationship between objects into a visually easy-to-understand form based on the similarity.

[0011] Thus, according to the evaluation method of the second invention, it is possible to actually perform multivariate analysis using multidimensional scaling, and to provide an index for evaluating general and complex information.

[0012] The evaluation method of the third invention is the evaluation method of the first invention, The object is a food or drink, and the characteristic is taste.

[0013] According to the evaluation method of the third invention, by treating the object as food or drink and the taste as the characteristic, it is possible to display as a map an evaluation index that clearly shows the characteristic quantities for the taste of food or drink, which has complex information.

[0014] In this way, according to the evaluation method of the third invention, when the object is food or drink, it is possible to provide an index for actually evaluating general and complex information regarding taste.

[0015] The evaluation method of the fourth invention is the evaluation method of the first invention, The object is a fermented food.

[0016] The evaluation method of the fifth invention is the fourth invention, The characteristic is a fermented state.

[0017] According to the evaluation methods of the fourth and fifth inventions, by using fermented foods as the target object, it is possible to display as a map an evaluation index that clearly shows the characteristic quantities for the fermentation state of food and beverages containing complex information.

[0018] In this way, according to the evaluation methods of the fourth and fifth inventions, when the subject is a fermented food, it is possible to provide an index for evaluating general and complex information regarding the actual fermentation state.

[0019] The evaluation method of the sixth invention is any one of the first to fifth inventions, Artificial intelligence is used in the mapping process.

[0020] According to the evaluation method of the sixth aspect of the invention, by using artificial intelligence (AI) such as machine learning in the mapping step, it becomes possible to easily realize advanced data analysis.

[0021] Thus, according to the evaluation method of the sixth aspect of the invention, it is possible to more easily and sophisticatedly analyze indices that evaluate general and complex information about an object.

[0022] The evaluation system of the seventh invention is an evaluation system for evaluating characteristics of an object based on a fluorescence spectrum, a data acquisition unit that acquires three-dimensional spectral data obtained by measuring the three-dimensional fluorescence spectrum of the object; a mapping unit that maps the results obtained by multivariate analysis of the three-dimensional spectral data into a two-dimensional space; The present invention is characterized by comprising:

[0023] According to the evaluation system of the seventh aspect of the invention, the mapping unit performs multivariate analysis on the three-dimensional spectrum data acquired by the data acquisition unit. Then, the mapping unit maps the results of the multivariate analysis onto two-dimensional space, so that the characteristics of the object, which exhibit a complex shape in the three-dimensional spectrum, can be displayed as a map in the two-dimensional space as evaluation indices with clear feature quantities as the results of the multivariate analysis.

[0024] In this way, the evaluation system of the seventh aspect of the invention can provide an index for evaluating general and complex information about an object. [Brief explanation of the drawings]

[0025] [Figure 1] FIG. 1 is a schematic diagram showing a configuration of an evaluation system according to an embodiment of the present invention. [Figure 2] This figure shows a contour map of the 3D fluorescence spectrum of sake, as part of the processing performed by the evaluation system in Figure 1. [Figure 3] FIG. 2 is a diagram showing two-dimensional mapping using multidimensional scaling analysis as part of the processing performed by the evaluation system shown in FIG. 1. [Figure 4] This figure shows the 3D fluorescence spectrum of a new sake, as part of the processing performed by the evaluation system in Figure 1. [Figure 5] This figure shows the time change in the 3D fluorescence spectrum of sake, as part of the processing performed by the evaluation system in Figure 1. [Figure 6] FIG. 2 shows the results of an analysis of the changes in sake over time as part of the processing performed by the evaluation system shown in FIG. 1. DETAILED DESCRIPTION OF THE INVENTION

[0026] An evaluation system according to one embodiment of the present invention will be described below with reference to FIG.

[0027] As shown in FIG. 1(A), the evaluation system 1 is an evaluation system for evaluating characteristics of an object based on the fluorescence spectrum, and includes a data acquisition unit 10 and a mapping unit 20.

[0028] The data acquisition unit 10 is a processing unit that executes the data acquisition step of the present invention, and acquires 3D fluorescence spectrum data obtained by measuring the three-dimensional fluorescence spectrum of the object X (hereinafter referred to as 3D fluorescence spectrum).

[0029] Furthermore, the data acquisition unit 10 includes a light emitting unit 11 that emits incident light of a predetermined wavelength toward the object X, and a light receiving unit 12 that receives fluorescence emitted from the object X. The light emitting unit 11 includes a light emitting controller (not shown) that controls the wavelength.

[0030] The mapping unit 20 is a processing unit that executes the mapping process of the present invention, and functions as an analysis unit that performs multivariate analysis of the 3D fluorescence spectrum data acquired by the data acquisition unit 10, and maps the results of the multivariate analysis into two-dimensional space.

[0031] Furthermore, the mapping unit 20 has a display unit 21 that displays the mapping results, and also has a database 22 for referencing the mapping results with past data.

[0032] The above is the configuration of the evaluation system 1 of this embodiment. In the above configuration, each of the processing units 10 to 22 is configured with hardware such as a CPU (Central Processing Unit), a ROM (Read Only Memory), and a RAM (Random Access Memory), and functions as an arithmetic unit (sequencer) for executing various processes by storing and holding programs for executing various processes in memory (not shown) and executing the programs. Also, some or all of the processing units 10 to 22 may be configured with other servers (external servers), etc., and the evaluation system 1 may be realized by distributed processing.

[0033] Next, the processing performed by the evaluation system 1 (the evaluation method of the present invention) will be described.

[0034] First, in the spectral data acquisition unit 10 of the evaluation system 1, incident light of a predetermined wavelength is incident on the object X from the light emitting unit 11, and the light receiving unit 12 acquires 3D fluorescence spectral data of the fluorescence emitted from the object X (corresponding to the data acquisition process of the present invention).

[0035] Next, in the mapping unit 20 of the evaluation system, the acquired 3D fluorescence spectrum is subjected to multivariate analysis using multidimensional scaling, and the results are mapped in two-dimensional space (corresponding to the mapping step of the present invention). [Example]

[0036] As Example 1, an embodiment will be described in which sake is used as the object X. Here, 50 types of sake (No. 1 to No. 50) that are generally available on the market were selected and their 3D fluorescence spectra were measured. 15 types were selected from the contour maps of the 3D fluorescence spectra and are shown in FIG. 2.

[0037] Here, the incident light was varied in the range of 320nm to 400nm, and the fluorescence emitted from sake when irradiated with the incident light was measured in the range of 350nm to 600nm. The incident light and fluorescence were set solely for this measurement, and are not limited to these wavelengths. Note that while 15 types of spectra are shown here, similar spectra were also measured for the remaining 35 types, and the results were plotted in 3D.

[0038] The 3D fluorescence spectrum diagrams are then used as data for multivariate analysis, with the incident light wavelength range and fluorescence measurement range set to the same range. Note that the incident light wavelength step and fluorescence measurement wavelength step are set to 1 nm during measurement, so the number of data points is the same for each diagram.

[0039] Next, to analyze the similarities between sake, this data was mapped onto two-dimensional space using multidimensional scaling analysis, a type of multivariate analysis. Figure 3 shows the results of the mapping. Images with similarities are placed close together. Furthermore, the X-axis value correlates with the amino acid content; the higher the amino acid content, the more mellow the sake, and the lower the amino acid content, the lighter the sake, confirming that it is possible to evaluate taste characteristics.

[0040] Additionally, a new type of sake was selected and its 3D fluorescence spectrum was measured. Figure 4 shows a contour map of the measured spectrum. This spectrum was analyzed using multidimensional scaling analysis, and by spotting it in the space shown in Figure 3, it was possible to evaluate the amino acid content and sakes with high similarity. [Example]

[0041] Next, as Example 2, another embodiment will be described in which sake is used as the object X. One type of sake (sk23: Junmai Daiginjo) that is commonly available on the market was selected and the 3D fluorescence spectrum was measured once a week. The measured contour map is shown in FIG.

[0042] Here, the incident light was varied between 320 and 400 nm, and the fluorescence emitted from sake when irradiated with the incident light was measured in the range of 350 to 600 nm. The incident light and fluorescence were set solely for this measurement and are not limited to these wavelengths. Note that while the spectra are shown for a one-week period, the period is not limited to this. Furthermore, the 3D spectral plots are used to generate data for multivariate analysis, with the incident light wavelength range and fluorescence measurement range being the same. Note that the incident wavelength step for the incident light wavelength and the measurement wavelength step for the fluorescence measurement were set to 1 nm during measurement, so the number of data points is the same for each plot.

[0043] To analyze the evaluation of sake maturation over time, these data were mapped onto two-dimensional space using multidimensional scaling analysis, a type of multivariate analysis. The mapping results are shown in Figure 6. Figure 6 also shows the results for other sakes (sk08, sk20, sk12; A to F are time series).

[0044] It was confirmed that the analytical values ​​of sk23 Junmai Daiginjo shifted upward to the right over time. When other sakes (sk08, sk20, sk12) were evaluated, some showed a downward shift to the right, some showed a circular change, and some showed almost no change.

[0045] From this, it was confirmed that by observing the trends of change over time, it is possible to classify sake into groups, evaluate the degree of maturation, and evaluate the state of chemical reactions within the sake.

[0046] Furthermore, even for the same brand of sake, the mapping position differs for special junmai sake (sk08), junmai daiginjo (sk08A), junmai sake (sk23), junmai daiginjo (sk23 open bottle), special junmai sake (sk20), and junmai ginjo (sk20A), allowing the rice polishing ratio and other factors to be evaluated.

[0047] As described above in detail, the evaluation method for evaluating the characteristics of the object X based on the fluorescence spectrum using the evaluation system 1 of this embodiment can provide an index for evaluating general and complex information about the object X.

[0048] In this embodiment, the mapping unit 20 may be configured to function as an analysis unit, and the evaluation of similarity by comparing with data stored in a database 22 that stores the results of past multivariate analyses may be performed using artificial intelligence (AI) such as machine learning.

[0049] Furthermore, in this embodiment, the object X has been described as being sake, a fermented food, but the object X is not limited to this and may be any food or drink (including pharmaceuticals), or any object other than food or drink that emits fluorescence.

[0050] Furthermore, if the object X is a food or beverage, the information about the food or beverage may be analyzed as characteristics, including some or all of the following: the taste of the food or beverage, the ingredients that make up the food or beverage, the shipping date of the food or beverage, and the packaging container of the food or beverage.

[0051] In this way, when the object X is food or drink, the state of the food or drink can be analyzed without relying on sensory evaluation or simple fluorescence measurement.

[0052] Furthermore, in this embodiment, multidimensional scaling has been used as an example of a multivariate analysis technique, but the multivariate analysis technique is not limited to this, and correspondence analysis including quantification type III may also be adopted. [Explanation of symbols]

[0053] 1...evaluation system, 10...data acquisition unit, 11...light emitting unit, 12...light receiving unit, 20...mapping unit, 21...display unit, 22...database, X...object.

Claims

1. An evaluation method for evaluating characteristics based on a fluorescence spectrum of an object, comprising: a data acquisition step of acquiring three-dimensional spectral data obtained by measuring the three-dimensional fluorescence spectrum of the object; a mapping step of mapping the results obtained by multivariate analysis of the three-dimensional spectral data onto a two-dimensional space; An evaluation method comprising:

2. 2. The evaluation method according to claim 1, wherein the multivariate analysis is a multidimensional scaling method.

3. 2. The evaluation method according to claim 1, wherein the object is a food or drink, and the characteristic is taste.

4. The evaluation method according to claim 1, wherein the object is a fermented food.

5. The evaluation method according to claim 4, wherein the characteristic is a fermentation state.

6. 6. The method of claim 1, wherein artificial intelligence is used in the mapping step.

7. An evaluation system for evaluating characteristics based on a fluorescence spectrum of an object, comprising: a data acquisition unit that acquires three-dimensional spectral data obtained by measuring the three-dimensional fluorescence spectrum of the object; a mapping unit that maps the results obtained by multivariate analysis of the three-dimensional spectral data into a two-dimensional space; An evaluation system comprising:

Images