KR101206295B1 - Discrimination of geographical origin of grain using near infrared multivariate analysis method - Google Patents
Discrimination of geographical origin of grain using near infrared multivariate analysis method Download PDFInfo
- Publication number
- KR101206295B1 KR101206295B1 KR20100041317A KR20100041317A KR101206295B1 KR 101206295 B1 KR101206295 B1 KR 101206295B1 KR 20100041317 A KR20100041317 A KR 20100041317A KR 20100041317 A KR20100041317 A KR 20100041317A KR 101206295 B1 KR101206295 B1 KR 101206295B1
- Authority
- KR
- South Korea
- Prior art keywords
- origin
- grain
- sample
- analysis
- samples
- Prior art date
Links
Images
Abstract
The present invention relates to a method for determining the origin of grain samples using near-infrared multivariate analysis, and more particularly, to investigate near-infrared rays in arbitrary amounts of grain samples to measure the near-infrared absorption spectra for each wavelength, and to develop and predict the origin discrimination prediction model. Obtain the raw absorption spectrum corrected by the average of the near-infrared absorption spectrum of the grain sample randomly divided for model verification and the development of the discriminant predictive model of origin of the grain sample, and the derivative thereof to check the difference in the spectrum of each origin. Develop a predictive model for determining the origin of grain samples by performing statistical analysis, and verify the suitability of the predictive model by comparing the developed predictive model of origin of the grain samples with the near-infrared absorption spectrum for verifying the predictive model of origin for the corresponding grain samples. And selecting an optimal model for predicting origin of origin It relates to a method for determining the origin of a grain sample comprising a.
According to the present invention, there is no need for a separate expensive device, and the time, cost, and effort required for analysis can be reduced. Even in the same case, it is possible to determine the country of origin according to the cultivation area with high accuracy, and to determine the country of origin in a short time of about 30 seconds per grain sample.
Description
The present invention relates to a method capable of determining the origin of various grain samples using near infrared irradiation and multivariate analysis.
According to the import and export of agricultural products, agricultural products from various countries of the world are introduced into Korea and are widely distributed. In particular, the problem is that foreign low-priced agricultural products or genetically modified agricultural products are sold domestically or are mixed with domestically sold products, which causes severe damage to domestic agricultural producers and provides relatively low-cost, low-quality foreign agricultural products. The damage to domestic consumers who know and consume expensive domestic agricultural products is also severe. Due to this situation, the national institution mandates the labeling of the country of origin of agricultural products and checks them from time to time.
The National Agricultural Products Quality Control Act (December 10, 2009) states, in Article 2, paragraph 7, “Geographic labeling means that agricultural products or their processed products, if the reputation, quality and other characteristics of the Refers to an indication that the processed product has been produced and processed in that particular area ”, and Article 2,
According to the revised law, even seeds of the same varieties with the same genes are domestic when grown in Korea and are cultivated in foreign countries.
On the other hand, most of the analysis methods currently used to determine the origin of domestic agricultural products are wet chemical analysis or extract method. In summary, quantitative DNA analysis (ginseng, rice, cabbage, red pepper, beef, etc.), X-ray fluorescence analysis (Barley, etc.), NMR spectroscopy (beans, sesame seeds, peony, creation, taxa, reeds, etc.), electronic nose analysis (Dangui, Cheongung, brown rice, ginseng, carrot, garlic, ganoderma, sesame seeds, sesame seeds, etc.), elemental analysis (barley) Etc.), Inductively Coupled Plasma Spectroscopy (Yanggi, etc.), Capillary Electrophoresis (Rice, Yulmu, Astragalus, Cheongung, etc.), Near Infrared Spectroscopy (Sesame, Sesame Oil, Astragalus, Green Tea, Pine, Astragalus, Safflower, Red Pepper, Ginseng) , Tobacco, missing, beef, lamb, flour, etc.) are mainly used.
Among these methods, PCR and RAPD methods, which are DNA analysis methods, are destructive methods and require qualitative PCR through sample crushing and pretreatment, DNA extraction and purification, DNA concentration conversion and purity measurement, and electrophoresis and quantitative PCR analysis. It is necessary to provide expensive equipment for a long time, long analysis period, primer set, fluorescence probe, standard material (CRM) or standard plasmid, etc. required for PCR. If the ability is necessary, and if the varieties with the same genetic information are cultivated in different countries, it is impossible to discriminate the area of crops at all.
In addition, X-ray fluorescence, NMR spectroscopy, electron nose analysis, elemental analysis, inductively coupled plasma spectroscopy, and capillary electrophoresis are also destructive analysis methods. The metabolites and their contents are extracted through crushing and pretreatment of crop samples of origin determination. As a wet chemical analysis method to be analyzed, expensive equipment, a long analysis time, various chemical reagents and skilled analysis techniques are required.
On the other hand, near-infrared spectroscopy uses near-infrared spectroscopy to irradiate near-infrared light in the 400 ~ 2,500nm region to a sample and analyze the reflected and transmitted spectra, which are reflected and transmitted. It is characterized by the absorption band of overtone and combination of fundamental frequencies of NH, OH and CH, and its strength is relatively weak. Until chemometrics were developed, they were difficult to distinguish from noise and were not used in actual analysis.However, since they were used by Karl Norris to analyze the moisture, fat and protein of agricultural products in agriculture, Its development is expanding its application to many fields.
This near-infrared spectroscopy reduces the time, cost, and effort required for analysis, eliminates the need for chemical reagents, and develops a variety of destructive or non-destructive techniques for the reuse and maintenance of analytical samples. It is widely used as a suitable method for characterizing crops and agricultural products in the state of containing.
However, to date, most of the country of origin determination using the near-infrared spectroscopy distinguishes the country of origin by simply comparing the spectrum of domestic and collected foreign products. In other words, in the case of simple collection, since the possibility of discrimination by different genetic background differences between domestic and foreign farm products is extremely high, it is not applicable to the determination of the origin of agricultural products with the same genetic background. However, there is no review on whether the seeds of domestically grown varieties with the same genetic information can be correctly identified even if they are grown in other countries.
In addition, most of the results of the determination of the origin by the near-infrared spectroscopy are the analysis of the powder state by the destructive method, in which the actual form of import and distribution of agricultural products is in a lost state, maintains original form, or some origin is When different agricultural products are mixed, in most cases, it is impossible to determine the country of origin.
Therefore, the inventors of the present invention minimize the burden of the existing method of determining the origin of agricultural products, such as the device cost, skilled technical skills and expertise of the analyst, manpower cost, and time required, and furthermore, the problem of being made in a destructive form. We tried to solve all the problems related to the determination of the origin of foreign cultivated agricultural products with the same genetic information as the cultivated genes, which can be the most serious error in discrimination. Small ring cups, horizontal manual modules, and rectangular cells, etc., are used to classify grain samples into various types of samples such as powder, granules, and populations. The resulting near-infrared absorption spectra are measured and their raw and derivatives obtained, and statistical analysis The present invention was completed by knowing that the country of origin can be determined by obtaining a difference according to the form, variety, strain and plantation of the grain sample.
According to the present invention, it is possible to precisely quantitatively and quantitatively analyze the origin of grain samples in powder, non-destructive form, or collectively, regardless of the shape of the grain sample, and in the case of grain samples having the same genetic information In the case of different cultivated lands, regardless of the distribution type of grains, it can be expected that the country of origin can be identified with an accuracy of 99%.
In addition, according to the present invention, any type such as powder, granules, and collective units in a grain sample can be expected to be able to quickly and accurately determine the origin in 30 seconds per analysis sample point in time.
In addition, according to the present invention, the burden of device cost, manpower, expenses and time is serious, it is possible to expect the advantages that can solve all the problems, such as the impossible to collect real-time analysis information.
As an example for achieving the above object, the grain sample origin determination method of the present invention, by irradiating near-infrared rays to a random amount of grain samples to measure the near-infrared absorption spectrum for each wavelength, and for the development of the prediction model of origin determination and prediction model verification Randomizing; Obtain the raw absorption spectrum corrected by the average of the near-infrared absorption spectra of the randomly divided grain sample and the derivative thereof for the development of a model for predicting the origin of the grain sample. Developing a prediction model of origin determination; And comparing the developed predictive model of origin determination of the grain sample with the near-infrared absorption spectrum for verifying the predictive model of origin determination of the corresponding grain sample to verify the suitability of the predictive model and to select an optimal predictive predictive model. Characterized in that made.
The statistical analysis of the derivatives includes Multiple Linear Regression (MLR), Partial Least Squares (PLS), Modified Partial Least Squares (MPLS), Principle Component Analysis (PCA). ) And Discriminant Analysis (DA).
Hereinafter, a method of determining the origin of grain samples using the near-infrared multivariate analysis method of the present invention will be described in detail for each step.
First, the near-infrared ray is irradiated to a random amount of grain samples to measure the near-infrared absorption spectra for each wavelength, and is randomly divided into a source predictive prediction model development and a prediction model verification.
The measurement of the near-infrared absorption spectrum can be performed differently according to the shape of the grain sample. When the grain sample is used as a powder, the near-infrared absorption is performed after being placed in a small ring cup as shown in FIG. 1. In the case of using a grain sample as a collective unit, it is preferable to carry out by irradiating near-infrared rays in a quartz container (rectangle cell) of quartz material as shown in FIG. As shown in FIG. 17, it is preferable to use a horizontal manual module that can measure grain samples individually by one grain in terms of measurement of grain samples.
The grain sample used in the present invention is sufficient as long as it can be eaten by humans or available to animals, and there is no particular limitation. Such grain samples may be of various varieties, strains and genetic information, and the same genetic information can be used to distinguish the country of origin from domestic and foreign because of different plantation. The varieties and tree numbers of grain samples may be adjusted differently according to the size of the near-infrared ray irradiator, but in terms of efficiency and accuracy of the predictive model, the varieties and tree numbers of samples may be adjusted to 100 to 500.
In the present invention, the near-infrared absorption spectrum measurement for determining the origin of grain samples is 400? It is desirable to measure the absorption spectrum between 2,500 nm. In general, the near infrared absorption spectrum is 700? Although absorption spectra between 2,500 nm are used, the absorption spectra may vary depending on the major components of the grain sample, which may provide important information for analysis of the origin of the grain sample. I can't.
Near-infrared absorption spectra of each measured wavelength of grain samples are arbitrarily divided into a source-determining predictive model development and a predictive model verification. For example, after measuring the near-infrared absorption spectra of 293 samples, the total 193 randomly selected out of 293 near-infrared absorption spectra are used for the development of predictive model for origin determination of grain samples, and the remaining 100 are used for verifying the predictive model for origin determination of grain samples. This study developed a number of predictive models suitable for determining the origin of grain samples using 193 absorption spectra, and the absorption spectra of 100 samples that were not used for predictive model development were selected predictive models suitable for determining the origin of developed grain samples. This is to verify whether the accuracy is high even when they are applied to the unknown sample.
Secondly, the raw absorption spectrum corrected by the average of the near-infrared absorption spectra of randomly divided grain samples and derivatives thereof were obtained for the purpose of developing a predictive model for determining the origin of the grain sample. This is the stage of developing a predictive model for determining the origin of grain samples.
In other words, the average raw near-infrared absorption spectrum, which is corrected by averaging the random-infrared absorption spectrum randomly divided for the development of the origin discriminant prediction model, by the number of grain samples used for analysis, is used as the raw near-infrared absorption spectrum and its derivative (first derivative). And second derivatives) to identify the differences in the spectrum of each country of origin, and to develop an optimal predictive model for determining the origin of grain samples by performing various statistical analyzes.
Methods applied to the statistical analysis include Multiple Linear Regression (MLR), Partial Least Squares (PLS), Modified Partial Least Squares (MPLS), Principal Component Analysis (PCA: Principle Component Analysis (Principle Component Analysis), Discriminant Analysis (DA) can be performed by any one of the methods, in particular, partial least squares (PLS), modified partial least squares (MPLS), principal component analysis (PCA), And multivariate analysis methods such as discriminant analysis (DA), the statistical analysis of the variables of an analyte sample included in a specific population without performing a full investigation of the chemical or physical characteristics of the sample Simultaneous analysis is possible, and the application of various multivariate methods differs from various wet and spectroscopic methods for determining origin. This is the advantage.
The measured near-infrared absorption spectra are absorbed shifts due to various factors. Therefore, when the first or second derivatives are treated with water, all near-infrared absorption spectra are aligned on the same baseline. This is to improve the discrimination ability by excluding the difference in absorbance derived from the phenomenon and using only the absorbance difference of specific absorption wavelengths, and the derivative processing makes the distinction and discrimination of a specific spectrum more clearly, which is highly recommended for analysis. Even the untrained public can easily determine the origin by eliminating the factors that disturb the analysis such as absorbance shift of the spectrum.
In addition, various statistical analyzes are performed because of the overlap of absorption wavelengths generated in the near infrared absorption spectrum. In most cases, because absorption wavelengths overlap, we want to develop the best prediction model by applying various statistical spectrum analysis methods to make more accurate predictions without the influence of overlapping wavelengths.
Thirdly, comparing the developed predictive model of origin determination of the grain sample with the near-infrared absorption spectrum for verifying the predictive model of origin determination of the corresponding grain sample to verify the suitability of the predictive model and select the optimal predictive predictive model. .
This is to analyze whether the origin prediction model of the developed grain sample is correctly applied to the unknown sample, and to select the optimal predictive model by measuring the precision at that time.
According to the present invention described above, by measuring the near-infrared absorption spectra and applying the multivariate analysis method, the origin of a grain sample can be easily determined by a non-transparent method, and in particular, the near-infrared absorption spectrum is determined regardless of the shape of the grain sample. Measurement and analysis can accurately determine the origin of a sample with only 30 seconds of analysis time per sample.
In addition, according to the present invention, it is possible to expect the effect of accurately determining the origin of the grain samples of the same genetic information, but the plantation is different, it is possible to provide a method for determining the country of origin that meets the geographic labeling standards in the Agricultural Products Quality Control Act.
In addition, according to the present invention, there is no need for a separate equipment used to determine the origin of the existing agricultural products, it is possible to shorten the analysis time required for a long time, and can expect the effect of improving the manpower, expenses and measurement errors.
In addition, when the present invention is carried out in a non-destructive method, it can be used for simultaneous analysis requiring rapidity, accuracy and real-time, there is an advantage that no pre-treatment operation that can reuse the grain sample after analysis.
In addition, the present invention is expected to provide powerful advantages that could not be imagined in the past in related industries such as the expansion of the application range of near-infrared spectroscopy technology and the important basic technology fields of device development.
Figure 1 is a photograph of applying a small ring cup (small ring cup) in the near infrared spectroscopy apparatus to measure the near infrared absorption spectrum of the powder sample in Example 1.
2 is an average raw near-infrared absorption spectrum of Korean and US soybean powder samples measured in Example 1 (red line: Korean, blue line: US).
Figure 3 is the spectrum of the average first derivative of the near-infrared absorption spectrum of the Korean and American soybean powder samples measured in Example 1 (red line: Korean, blue line: US).
Figure 4 is the spectrum of the average second derivative of the near-infrared absorption spectrum of Korean and US soybean powder samples measured in Example 1 (red line: Korean, blue line: US).
5 is a graph showing three-dimensional three-dimensional discrimination pattern according to the principal component analysis of the near-infrared absorption spectrum of Korean and US soybean powder samples measured in Example 1 [+: Korean, □: American].
FIG. 6 is a graph showing a discrimination pattern by partial least-squares discrimination analysis of near-infrared absorption spectra of Korean and US soybean powder samples measured in Example 1 [+: Korean, □: American].
Figure 7 is a graph showing the discrimination pattern by the modified partial least-squares analysis of the near-infrared absorption spectrum of Korean and American soybean powder samples measured in Example 1.
FIG. 8 is a graph showing an unknown sample discrimination pattern by modified partial least squares analysis of near-infrared absorption spectra of Korean and US soybean powder samples measured in Example 1. FIG.
FIG. 9 is a photograph of applying a rectangular cell in a near-infrared spectrometer to measure the near-infrared absorption spectrum of a collective seed sample in Example 2. FIG.
10 is an average raw near-infrared absorption spectrum of Korean and US soybean seed samples measured in Example 2 (red line: Korean, blue line: US).
FIG. 11 is the mean first derivative spectrum of the near infrared absorption spectrum of Korean and US soybean seed samples measured in Example 2 [red line: Korean, blue line: American].
12 is an average second derivative spectrum of the near-infrared absorption spectra of Korean and US soybean seed samples measured in Example 2 (red line: Korean, blue line: American).
FIG. 13 is a graph showing three-dimensional three-dimensional discrimination patterns according to principal component analysis of near-infrared absorption spectra of Korean and US soybean seed samples measured in Example 2 [+: Korean, □: American].
FIG. 14 is a graph showing a discrimination pattern by partial least-squares discrimination analysis of near-infrared absorption spectra of Korean and US soybean seed samples measured in Example 2 [+: Korean, □: American].
Figure 15 is a graph showing the discrimination pattern by the modified partial least-squares analysis of the near-infrared absorption spectrum of the Korean and American soybean seed samples measured in Example 2.
Figure 16 is a graph showing the unknown pattern discrimination pattern by the modified partial least-squares analysis of the near-infrared absorption spectrum of the Korean and American soybean seed samples measured in Example 2.
FIG. 17 is a photograph to which a horizontal manual measurement apparatus is applied in a near infrared spectroscopy apparatus to measure a near infrared absorption spectrum of a grain seed single grain sample in Example 3. FIG.
18 is an average raw near-infrared absorption spectrum of a sample of Korean and US soybeans measured in Example 3 (red line: Korean, blue line: American).
19 is an average first derivative spectrum of the near-infrared absorption spectrum of Korean and US soybean grain samples measured in Example 3 (red line: Korean, blue line: US).
20 is an average second derivative spectrum of the near-infrared absorption spectrum of Korean and US soybean grain samples measured in Example 3 (red line: Korean, blue line: US).
Figure 21 is a graph showing the three-dimensional three-dimensional discrimination pattern according to the principal component analysis of the near-infrared absorption spectrum of Korean and
FIG. 22 is a graph showing a discrimination pattern by partial least-squares discrimination analysis of near-infrared absorption spectra of Korean and
FIG. 23 is a graph showing a discrimination pattern by modified partial least squares analysis of near-infrared absorption spectra of Korean and
24 is a graph showing the unknown sample discrimination pattern by the modified partial least-squares analysis of the near-infrared absorption spectrum of Korean and
EMBODIMENT OF THE INVENTION Hereinafter, this invention is demonstrated in detail based on an Example, drawing, etc. However, what is described in the following examples and drawings is only one example presented for carrying out the present invention, and the scope of the present invention is not limited by the following examples and drawings.
Example 1. Development of Prediction Model for Determination of Soybean Origin for Powder Samples
1) Development of predictive model
Soybean used in the embodiment of the present invention is a domestically grown soybean varieties and domestic and foreign soybean gene source 293 strains having a variety of genetic information, these cultivated and harvested in domestic redistribution was evaluated in Korea, these 293 strains harvest In the following year, more than 100 seeds of each strain were grown and harvested in the same way in the United States, and used as an analytical material for determining the origin. The near-infrared spectrometer used was NIRS 6500 of Foss (U.S.A).
Soybean samples from 293 Korean and US soybeans harvested from each region were ground with a micro hammer mill (manufactured by Culatty AG, Zurich, Switzerland), filtered using a 1.0 mm screen, and sealed in a sealed plastic bag at 4 ° C until use. Stored.
Powder samples of Korean and American soybeans, each of 293 series, were measured between 400 and 2500 nm using a transport device using a small ring cup in a near-infrared spectrometer having a structure as shown in FIG. The absorbance spectra of each wavelength were measured to obtain the raw near-infrared absorption spectra. The measured 293 Korean powders and 293 American powders were randomly divided into two sets.
That is, one set of each Korean and US powder sample is used as a sample for developing the origin discrimination prediction model using each 193 system, and the other set of the origin discrimination prediction model developed using each Korean and 100 American origin The goodness-of-fit of the origin discriminant predictive model (calibration formula) developed using the external validation set (for predictive model verification) for unknown sample evaluation was tested.
The raw near-infrared absorption spectra of each sample of 193 strains for developing Korean and US origin predictive models were averaged by the number of measured soybean strains (193), and the average near-infrared absorption spectra of each Korean and American soybean powder sample. It is shown in Figure 2 separately. As a result, there is a difference in the near-infrared absorption spectra of Korean and American soybean powder samples in some wavelength ranges of 400 ~ 2,500nm, so that even if the soybeans with the same genetic information have different origins, the near-infrared absorption characteristics will be different. I could confirm it.
In addition, the near-infrared absorption spectrum characteristics of the average raw spectrum of Korean and US soybean powder samples having the same genetic information measured at all wavelengths between 400 and 2500 nm when the first and second derivatives are applied are shown in FIGS. 3 and 4. Shown. As a result, when the first derivative was applied, the average spectrum of the soybean powder sample originating in the United States was slightly lower than that of the Korean soybean powder sample at 1,890 ~ 2,040nm, while the average spectrum of the Korean soybean powder sample was At 686 nm, the absorbance characteristics were slightly lower than those of US soybean powder samples, so that the soybean powder samples of different origins from Korea and the US could be distinguished due to the difference in absorption spectrum in the specific wavelength range. In addition, even when the second derivative shown in FIG. 4 was applied, the average spectrum of Korean and US soybean powder samples showed a difference in absorption spectrum at 656, 685, 1368, 1458, 1482, and 1906 nm, and the result of the first derivative shown in FIG. Similarly, the origin of Korean and American soybean powder samples could be determined by the difference of absorption spectrum.
On the other hand, using the near-infrared absorption spectrum of soybean powder samples of different origin, the results of examining the three-dimensional three-dimensional discrimination between Korean and US soybean powder samples through the principal component analysis (PCA: Principle Component Analysis) is shown in FIG. As a result, Korean and American soybean powder samples clearly showed a difference in three-dimensional space. Most of Korean soybean powder samples were distributed in 1, 2, and 3D planes. In this case, it is distributed in the rear in the 2D plane and in the upper right in the 3D plane, and it can be clearly seen that the distinction between the groups of Korean and US soybean powder samples can be clearly distinguished.
In addition, when the second derivative is applied to the raw absorption spectra of soybean powders of each origin, it is made in Korea using PLSDA (Partial Least Squares Discriminant Analysis), which is one of the multivariate statistical analysis methods of near-infrared spectroscopy. And the result of examining the discriminating power of the US soybean powder sample is shown in FIG. In the partial least squares analysis, the variable to be discriminated is calculated by dividing it into 1 or 2 by two classification methods. When the US soybean group is divided into 1 and the Korean soybean powder sample is divided into 2, As a result of the analysis, the scattering distribution of the US soybean powder sample was less than 1.3 and the Korean soybean was 1.6 or more. Therefore, it is possible to accurately distinguish between Korean and US soybean powder samples by partial least-squares discrimination analysis. Derived.
In the above results, when the same plant was produced in 293 different fields, the raw spectrum, primary and secondary derivative spectra of soybean powder samples of each origin, and principal component analysis, which is a kind of statistical multivariate analysis, Country of origin was obtained by dimensional stereoscopic analysis and partial least-squares discrimination analysis.
However, in the case of daily routine analysis, highly skilled knowledge is needed to interpret the origin by only spectrum, so it is necessary to develop a predictive model of origin discrimination that can be routinely analyzed by statistical analysis of spectrum. The first and second derivatives were applied to the raw near-infrared absorption spectra of, and the selection of the predictive model of origin discrimination using the partial least squares analysis was performed statistically and the results are shown in Table 1 below.
Determined number
Can't
Uncertain number
(193 system)
(193 system)
In the results of Table 1, when the partial least-squares discriminant analysis method is applied, the discriminant predictive model using the raw, first and second derivatives shows 100% accuracy of Korean and US soybean powder samples. It can be determined that can be determined.
2) Prediction Model Verification
The applicability of the unknown sample was verified to evaluate whether the predictive model developed using the partial least square analysis method can be actually used as a predictive model for origin determination when the routine analysis for the origin determination of soybean powder samples is applied.
The test materials were reviewed for applicability of unknown samples for origin determination using 100 sets of external verification sets prepared for evaluation of unknown samples of origin discriminant predictive models developed among 293 Korean and US soybean powder samples. By evaluating the accuracy of the origin determination by applying the developed partial discriminant discrimination analysis method according to various derivatives, the present invention can be used for discriminating Korean and US soybean powder samples by routine methods. Review.
Table 2 shows the results of verifying the accuracy when using the powder sample and applying the developed least predictive analysis predictive model for discriminating the origin of Korean and US soybeans to the unknown sample.
Determined number
Can't
Uncertain number
(100 lines)
(100 lines)
As can be seen from the results of Table 2, in the second derivative of the partial least-squares discriminant analysis predictive model for origin determination applicable to soybean powder samples, the discrimination accuracy of Korean products is 98%, and the discrimination accuracy of US products is 100%. In the case of soybean powder samples, it showed reliable superior origin discrimination accuracy in partial least-squares discrimination analysis using second derivative.
3) Development of prediction model and verification of prediction model by applying other multivariate analysis methods
On the other hand, various derivatives are applied to the raw absorption spectrum of soybean powder samples in order to examine whether it is possible to discriminate by other multivariate analysis methods in addition to the partial least-squares discriminant analysis method of origin discrimination prediction model using multivariate analysis method. Table 3 shows the results of examining the predictive model for determining the origin of Korean and US soybean powder samples using Modified Partial Least Squares (MPLS), one of the statistical methods.
treatment
In the modified partial least-squares method, 193 spectra of different origins were mixed with each other in the software, and the variables were set by dividing the variables to be discriminated into two at random.The Korean powder sample was 1 and the US powder sample was 100. Multivariate analysis was performed using the modified partial least squares method.
As shown in Table 3 above, when the variable of Korean soybean powder sample is set to 1 and the value of US soybean powder sample is set to 100, the result of determination of origin according to the modified partial least-squares method is determined by applying the second derivative ( R 2 ) was 0.996 to determine the origin of high significance (Fig. 7), the scattering distribution is accurately divided between the random variable -10 ~ 15 range in the case of Korea, between the random variable 90-110 in the case of the United States As a result, the modified partial least-squares analysis can also be used to obtain a result that discrimination between Korean and US soybean powder samples is possible at a highly accurate level.
In addition, the results of verifying the accuracy when applied to 100 unknown samples from Korea and the United States in order to examine the applicability of the unknown model using the modified partial least-squares analysis method for the optimized powder sample are shown in Table 4 below. It was.
Bias: average difference between reference and NIRS values.
r2: coefficient of determination of cross-validation. SEP (C): the corrected standard error of prediction. RSP: SD / SEP (C), the ratio of SD of reference data to SEP (C) in the external validation set.
As can be seen from the results of Table 4, the modified partial least-squares predictive model developed for origin determination using the second derivative has a coefficient of determination ( r 2 ) of 0.993, which is similar to that of development of the predictive model. The randomness of the sample was represented between -15 and 12 in Korea and 90 to 110 in the United States, respectively, and showed a significant difference based on 50, the boundary of the random variable for origin determination. In this case, it can be seen that the superior origin can be reliably determined even by the modified partial least-squares analysis using the second derivative (FIG. 8).
Example 2. Nondestructive Development of Predictive Model for Origin Discrimination for Collective Samples
1) Development of predictive model
The test material was the same as that used in Example 1, and was only intended to determine the origin by non-destructively analyzing the seed itself without pulverizing the soybean sample. The near-infrared spectrometer used was NIRS 6500 of Foss (U.S.A).
For the 293 strains of Korean and US soybeans, each wavelength between 400 and 2,500 nm was used in a near-infrared spectrometer having a structure as shown in FIG. 9 using a transport apparatus using a rectangular cell. Absorbance spectra were measured to determine the raw near-infrared absorption spectra. The 293 Korean and 293 American strains of each seed state were randomly divided into two sets. In other words, one set of each Korean and US seed sample is used as a sample for developing a predictive model of origin determination using each 193 system, and the other set is unknown of the origin discrimination predictive model developed using 100 Korean and US origins. The goodness-of-fit of the origin discriminant predictive model (calibration formula) developed using the external validation set (for predictive model validation) for sample evaluation was tested.
The raw near-infrared absorption spectra of each sample of 193 strains for the development of Korean and US origin predictive prediction models were averaged by the number of measured soybean strains (193), respectively. 10 is shown. As a result, there is a difference in absorption spectra of Korean and US soybean seed materials in some wavelength ranges of all 400 ~ 2,500nm wavelength ranges. It could be seen that the external absorption characteristics are different from each other.
In addition, the average raw near-infrared absorption spectra of each Korean and US soybean seed material with the same genetic information measured at all wavelengths between 400 and 2500 nm are plotted with the near-infrared absorption spectrum when the first and second derivatives are applied. Shown in 11 and 12. As a result, when the first derivative was applied, the average spectrum of the soybean seed sample originating in the United States was similar to that of the soybean powder sample of Example 1, showing a characteristic light absorbency slightly lower than that of the Korean soybean seed sample at 1,890? 2,040 nm. On the other hand, the average spectrum of Korean soybean powder samples showed a slightly higher absorbance than that of US soybean powder samples at 688 nm, so it was possible to distinguish soybean seeds from different origins from Korea and the United States due to the difference in absorption spectrum in the specific wavelength range. . In addition, even when applying the second derivative shown in Figure 12, the average spectrum of Korean and US soybean seed samples was 626, 684, 764, 962, 1148, 1400, 1512, 1672, 1894, 1964, 2044, 2358, and 2450 nm. The difference between the absorption spectrum and the origin of Korean and US soybean seeds was possible.
In addition, Fig. 13 shows the results of examining the discriminating power in three-dimensional stereoscopic spaces of Korean and US soybean seed samples through principal component analysis, which is one of the multivariate analysis methods, using near-infrared absorption spectra of soybean seed materials of different origin. It was. In the three-dimensional space, the location of Korean and American soybean seedlings was clearly different. Most of the Korean soybean seedlings were distributed in 1, 2, and 3D planes. In the case of the two-dimensional plane, it was distributed behind the Korean and the upper right in the three-dimensional plane. Showed a dense pattern.
On the other hand, when the second derivative is applied to the raw near-infrared absorption spectrum of soybean seed by each origin, the discrimination ability of Korean and US soybean seed material is determined by using partial least-squares discrimination analysis, which is one of the multivariate statistical analysis methods of near-infrared spectral spectrum. Examination results are shown in FIG. In the partial least squares analysis, the variable to be discriminated is calculated by dividing it into 1 or 2 by two classification methods. When the US soybean group is divided into 1 and the Korean soybean powder sample is divided into 2, As a result of the analysis, it was confirmed that the US soybeans were less than 1.3 and the Korean soybean powder samples were scattered more than 1.5. .
In the above results, when the same plant was produced in 293 different fields, the raw spectrum, first and second derivative spectra of soybean seed material of each origin, and principal component analysis, which is a kind of statistical multivariate analysis method, were classified. Country of origin can be determined by dimensional stereoscopic analysis and partial least-squares discriminant analysis method.In case of routine analysis, highly skilled knowledge is needed to interpret the country of origin only by spectrum. Since it is necessary to develop a multivariate predictive model for discrimination, it is necessary to apply the first and second derivatives to the raw near-infrared absorption spectra of Korean and American soybean seeds, and to select the multivariate predictive model using the partial least-squares discriminant analysis. Was performed statistically, and the result was Table 5 shows.
(193 system)
(193 system)
In the results of Table 5, when applying the partial least-squares discriminant analysis of the statistical multivariate analysis, in the application of the raw and first derivatives, the prediction model developed in both the Korean and the American soybean seed samples showed 100% accuracy, and thus the perfect origin can be determined. In the origin discriminant prediction model developed by applying the second derivative, it was confirmed that 1 line of the sample was indeterminate in Korean soybean seed samples among 193 samples, and the accuracy was 99.5%. When the partial least-squares discrimination analysis was applied, it was confirmed that there was at least 99.5% of origin discrimination ability under most derivative conditions.
2) Prediction Model Verification
The applicability of unknown samples was verified to evaluate whether the predictive models for each derivative condition developed using the partial least-squares discrimination method can be used for the determination of the origin when the daily analysis for the origin of soybean seed is applied. . The test materials were reviewed for applicability of unknown sample to origin determination using 100 sets of external verification set prepared for the unknown sample evaluation of origin discriminant predictive model developed among 293 Korean and American soybean seed materials. By re-evaluating the accuracy of the origin discrimination by applying the developed partial discriminant discrimination analysis method according to the various derivatives according to the above-described derivatives, whether the present invention is usable for discriminating Korean and US soybean seed materials by the routine method. Was reviewed.
Table 6 shows the results of verifying the accuracy when the soybean seed material was used and the developed partial least-squares discriminant analysis predictive model for Korean and US soybean origin was applied to the unknown sample.
Determined number
Can't
Uncertain number
(100 lines)
(100 lines)
In the results of Table 6, the second derivative of the partial least-squares discriminant analysis predictive model that can be applied to soybean seed material shows 98% accuracy in Korea and 100% in US. In addition, the partial least-squares discriminant analysis using the second derivative also showed the superior accuracy of the origin discrimination accuracy.
3) Development of prediction model and verification of prediction model by applying other multivariate analysis methods
In addition, various derivatives are applied to the raw near-infrared absorption spectrum of the seed material in order to examine whether the origin of the bean seed material can be determined by other multivariate analysis methods in addition to the partial least-squares discrimination method among the multivariate analysis prediction models. Table 7 shows the results of examining the predictive model for determining the origin of Korean and US soybean seed using the modified partial least square method, which is one of the multivariate statistical analysis.
treatment
In the modified partial least-squares method, 193 spectrums of different origins were mixed with each other in software, and the variables were set by dividing the variables to be discriminated into two at random.The Korean termination fee was 1 and the American termination fee was 100. Multivariate analysis was performed using the modified partial least squares method. When the Korean soybean seed material was set to 1 and the US soybean seed material was set to 100, the determination coefficient of origin according to the modified partial least-squares method showed that the coefficient of determination ( R 2 ) was 0.996 when the first derivative was applied. It was possible to determine the origin where the significance was secured (Fig. 7). In Korea, the scattering distribution is accurately classified between the random variable range of -10 to 13 and the US variable between the random variable 87 to 111. Square analysis also showed that the origin of Korean and US soybean seeds could be identified at a highly accurate level.
Table 8 shows the results of verifying the accuracy when applied to 100 unknown samples from Korea and the United States in order to examine the applicability of unknown samples using the modified partial least-squares analysis method for optimal soybean seeds. Indicated.
In the results of Table 8, the modified partial least-squares predictive model developed for the origin discriminant predictive model using the first derivative has a decision coefficient ( r 2 ) of 0.993, which shows a level of discrimination accuracy similar to that in developing the predictive model. The random variables represent between -8 and 14 in Korea and between 90 and 112 in the United States, respectively, and show a significant difference based on 50, which is the boundary of the random variable for origin determination. It was also found that the superior origin can be reliably determined by the modified partial least-squares analysis using (Fig. 16).
Example
3.
1) Development of predictive model
As the test material, the same material as that used in Example 2 was used, and only one grain of soybean was analyzed nondestructively to determine the origin. The near-infrared spectrometer used was NIRS 6500 of Foss (U.S.A).
Samples of 193 units of 293 strains of Korean and American soybeans were selected between 400 and 2,500 nm by selecting 50 grains from each strain of Korean and American strains using a horizontal manual measurement device in a near-infrared spectrometer as shown in FIG. The spectrum was measured in the near-infrared region, and the measured near-infrared absorption spectra of each of the strains were averaged again to correct the representative near-infrared spectra for each strain, and the same method was used for the 293 strains for both Korean and American strains. It was used after calibration with measurement and average spectrum.
After the measurement, the Korean 293 strains and the US 293 strains, which were averaged to each grain state, were randomly divided into two sets. In other words, one set of each Korean and one US sample is used as a sample for developing a predictive model for origin determination using each 193 system, and the other set is a multivariate analysis prediction for origin discrimination developed using each Korean and 100 American products. The suitability of the developed predictive model (calibration formula) for origin determination was tested using the external verification set (for predicting predictive model) for evaluation of unknown samples.
The raw near-infrared absorption spectra of 193 samples for each Korean and US origin predictive model development were again averaged by the number of measured soybean strains (193), and the average raw values of each Korean and American soybean sample. The near-infrared absorption spectrum is shown in FIG. 18 separately. As a result, the absorption spectra of Korean and American soybean samples were obviously different in some wavelength ranges from 400 to 2,500 nm, so that even if the soybeans with the same genetic information were grown in different origins, the NIR absorption characteristics of one grain unit would be different. It can be seen that these appear differently. In addition, the average raw near-infrared absorption spectra of Korean and US soybean samples with the same genetic information measured at all wavelengths between 400 and 2,500 nm are shown in the near-infrared absorption spectrum when the first and second derivatives are applied. Shown at 19 and 20. As a result, when the first derivative was applied, the absorption pattern of the spectrum was almost the same shape according to the difference of origin, but there was a slight difference in absorbance according to the wavelength region. In addition, the average near-infrared absorption spectrum of Korean and American soybean samples was similar when the second derivative was shown in FIG. 20, and the absorbance characteristics were different at the characteristic wavelengths. Different aspects were identified.
Fig. 21 shows the result of examining the discriminating power in three-dimensional three-dimensional structure of Korean and US soybean samples by using principal component analysis (PCA) using near-infrared absorption spectra of one soybean sample having different origins. As a result, one sample of Korean and US soybeans clearly showed a difference in three-dimensional space, and most of the samples of Korean soybeans were distributed in the middle of one, two, and three-dimensional planes. In the case of the lip sample, it is distributed at the bottom, and it can be clearly seen that the distinction between the groups of Korean and US soybean samples can be distinguished.
In addition, when the second derivative is applied to the raw absorption spectrum of one soybean sample of each origin, it is possible to discriminate between Korean and US soybean samples using partial least-squares discriminant analysis, which is one of the multivariate statistical analysis of near-infrared spectroscopy. The results of examining the results are shown in FIG. 22. In the partial least squares analysis, the variable to be discriminated is calculated by dividing it into 1 or 2 by two classification methods. When the US soybean group is divided into 1 and the Korean soybean powder sample is divided into 2, As a result of the analysis, it can be seen that the scattered distribution of the US soybean samples is less than 1.2 and that the Korean soybeans are more than 1.6, so that the discrimination between Korean and US soybean samples can be accurately determined by the partial least-squares discrimination analysis. The results were derived.
From these results, the same 293 strains were produced in Korea and the United States, with different plantations produced, based on the raw near-infrared absorption spectra, primary and secondary derivative spectra of one soybean sample, and primary and secondary variance analysis. 3D three-dimensional analysis and partial least-squares discriminant analysis result showed that the country of origin can be discriminated, and in case of routine analysis, highly skilled knowledge is needed to interpret the country of origin only by spectrum. It is necessary to develop a predictive model of origin discrimination that can be routinely analyzed by using the first and second derivatives in the raw near-infrared absorption spectra of Korean and American soybean samples. Multivariate analysis using the partial least-squares discrimination analysis The following table shows the results of statistically selecting the predictive model. 9 is shown.
Determined number
Can't
Uncertain number
(193 system)
(193 system)
In the results of Table 9, when the partial least-squares discriminant analysis is applied, the discriminant accuracy of Korean and American soybean samples is 99.0 to 100%. It was found that the country of origin can be almost completely identified, and in particular, it was confirmed that both Korean and US were 100% accurate when the second derivative was applied.
2) Prediction Model Verification
The applicability of unknown samples was verified to evaluate whether the origin discriminant predictive model developed using the partial least-squares discriminant analysis method can be actually used for the origin discrimination when the routine analysis for the origin determination of one soybean sample is applied.
The test material was reviewed for applicability of unknown sample to origin determination using 100 sets of external verification set prepared for evaluation of unknown sample of developed origin discriminant predictive model among 293 Korean and US soybean powder samples. The present invention is used for discriminating Korean and US soybean samples by routine method by evaluating the accuracy of origin discrimination by applying to the discriminant prediction model of origin using the various partial least-squares discriminant analysis methods developed above. Review if possible.
Table 1 shows the results of verifying the accuracy when using the sample of one lip unit and applying the developed least-squares prediction analysis model for discriminating the origin of Korean and US soybeans.
Determined number
Can't
Uncertain number
(100 lines)
(100 lines)
As can be seen in Table 10, in the case of application of the second derivative of the partial least-squares discriminant analysis predictive model for the determination of origin applicable to one soybean grain sample, the discrimination accuracy of Korean products is 97% and that of the USA is 100%. Although the evaluation accuracy of Korean products is slightly lower than that of soybean powder and seed material, it showed excellent origin determination accuracy that was reliable in origin discrimination analysis.
3) Development of prediction model and verification of prediction model by applying other multivariate analysis methods
In addition, various derivatives are applied to the raw near-infrared absorption spectrum of a single soybean sample to examine whether it can be discriminated by other multivariate analysis methods in addition to the partial least-squares discriminant analysis model among multivariate analysis prediction models. Table 11 shows the results of examining the predictive model for determining the origin of Korean and US soybean samples using the modified partial least-squares method, one of the analysis methods.
treatment
In the modified partial least-squares method, 193 spectra of different origins were mixed with each other in software, and the variables were set by dividing each variable into two randomly. A multivariate analysis was performed using the modified partial least squares method.
As shown in Table 11 above, when the variable of one Korean soybean sample was set to 1 and the variable of one soybean sample from United States was set to 100, the determination result of origin determination according to the modified partial least squares method was determined when the second derivative was applied. The origin ( R 2 ) was 0.999 with high significance, and it was possible to determine the origin (Fig. 23). Since the distribution can be confirmed, it was confirmed that the origin of Korean and US soybean samples can be accurately determined by the modified partial least squares analysis.
In addition, in order to examine the applicability of the unknown sample of the origin discriminant prediction model using the modified partial least squares analysis method for the optimal one-piece sample, the results of verifying the accuracy when applied to 100 unknown samples of Korean and US samples are as follows. Table 12 shows.
As shown in Table 12, the modified partial least-squares predictive model developed for the determination of the origin using the second derivative has a coefficient of determination ( r 2 ) of 0.999, which has the same level of discrimination accuracy as that of the development of the predictive model. Each random variable represents between -4? 5 for Korean and 95 ~ 105 for US, respectively, and shows a significant difference based on 50, which is the boundary of the random variable for origin determination. It was confirmed again that there was a reliable superior origin discrimination ability by the modified partial least-squares analysis using the second derivative (FIG. 24).
The present invention described above is not limited to the above-described embodiment and the accompanying drawings, and various substitutions, modifications, and changes are possible within the scope without departing from the technical spirit of the present invention. It will be evident to those who have knowledge of.
In each figure, the red line represents the result from Korea, and the blue line represents the result from the United States. In addition, (+) indicates a result from Korea, and (□) indicates a result from the United States.
Claims (7)
S1) measuring near-infrared absorption spectra for each wavelength by irradiating near-infrared light on a grain, a grain seed grain or a grain sample in a collective grain seed state, and randomly classifying it for development of a prediction model for origin determination and verification of a prediction model;
S2) Obtain the raw absorption spectrum corrected by the average of the near-infrared absorption spectra of the randomly divided grain samples and the derivative thereof for developing the discriminant predictive model of origin of the grain sample, and confirm the difference between the spectrums of each country of origin and perform statistical analysis. Developing a predictive model for determining the origin of the sample; And
S3) verifying the suitability of the predictive model and selecting an optimal origin discrimination prediction model by comparing the developed predictive model of origin determination of the grain sample with the near-infrared absorption spectrum for verifying the predictive model of origin discrimination of the corresponding grain sample;
Country of origin determination method of a grain sample comprising a.
The grain sample is the origin of the grain sample determination method, characterized in that consisting of a sample consisting of a variety of varieties, lines, and genetic information.
Wherein the grain sample has the same genetic information and the plantation determination method of origin of the grain sample, characterized in that consisting of different.
The near-infrared absorption spectrum is a determination method of origin of grain samples, characterized in that using the absorption spectrum in the wavelength range of 400 to 2,500nm.
In addition to the near-infrared absorption spectrum, the origin of the grain sample is measured using any one selected from infrared (IR) spectroscopy, ultraviolet-vis spectroscopy, and ROMAN spectroscopy. Discrimination method.
The statistical analysis of the derivatives includes Multiple Linear Regression (MLR), Partial Least Squares (PLS), Modified Partial Least Squares (MPLS), Principle Component Analysis (PCA). And a discriminant analysis (DA) method for determining the country of origin of farm products.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR20100041317A KR101206295B1 (en) | 2010-05-03 | 2010-05-03 | Discrimination of geographical origin of grain using near infrared multivariate analysis method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR20100041317A KR101206295B1 (en) | 2010-05-03 | 2010-05-03 | Discrimination of geographical origin of grain using near infrared multivariate analysis method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| KR20110121838A KR20110121838A (en) | 2011-11-09 |
| KR101206295B1 true KR101206295B1 (en) | 2012-11-29 |
Family
ID=45392562
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| KR20100041317A KR101206295B1 (en) | 2010-05-03 | 2010-05-03 | Discrimination of geographical origin of grain using near infrared multivariate analysis method |
Country Status (1)
| Country | Link |
|---|---|
| KR (1) | KR101206295B1 (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR102291771B1 (en) | 2020-06-23 | 2021-08-19 | 강원대학교산학협력단 | Non-destructive Determination Method For Plant Seed Germination And Growth Using Near-Infrared Absorption Spectrum |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101721735B1 (en) * | 2014-09-29 | 2017-03-30 | 경북대학교 산학협력단 | Rapid discrimination of F1 hybrid seeds from their parental lines using FT-IR spectroscopy combined by multivariate analysis |
| KR101660482B1 (en) * | 2014-11-28 | 2016-09-29 | 대한민국 | Methods for Discriminating the Origins of Red Ginseng Extract Using Near infrared reflectance Spectroscopy |
| KR101676125B1 (en) * | 2014-12-12 | 2016-11-14 | 경북대학교 산학협력단 | Discrimination Method of Cultivar and Origin from Sepal of Dry Persimmon Using FT-IR Spectroscopy Combined by Multivariate Analysis |
| CN106372342A (en) * | 2016-09-05 | 2017-02-01 | 中山大学 | Design method of higher-order digital differentiator based on genetic algorithm |
| CN106560698A (en) * | 2016-10-20 | 2017-04-12 | 中国计量大学 | Identification method for producing area of plant based on multiple detection technologies |
| CN109975236A (en) * | 2019-04-04 | 2019-07-05 | 山东省农业科学院农业质量标准与检测技术研究所 | A method of identifying the honeysuckle place of production using near-infrared spectrum technique |
| CN113092407B (en) * | 2021-04-20 | 2024-01-30 | 深圳华创智信科技有限公司 | Near infrared spectrum-based method for identifying varieties of original places of five-normal rice flower fragrance rice |
| CN117647501B (en) * | 2024-01-29 | 2024-04-26 | 中国计量科学研究院 | ICP spectral feature-based intelligent detection method and system for elements in food |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100334497B1 (en) | 1999-05-17 | 2002-04-26 | 김효진 | Apparatus for analysing spectrom of portable near-infrared |
| KR100883664B1 (en) | 2008-11-07 | 2009-02-18 | 강원대학교산학협력단 | Non-destructive discriminant analysis of genetically modified crops |
-
2010
- 2010-05-03 KR KR20100041317A patent/KR101206295B1/en not_active IP Right Cessation
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100334497B1 (en) | 1999-05-17 | 2002-04-26 | 김효진 | Apparatus for analysing spectrom of portable near-infrared |
| KR100883664B1 (en) | 2008-11-07 | 2009-02-18 | 강원대학교산학협력단 | Non-destructive discriminant analysis of genetically modified crops |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR102291771B1 (en) | 2020-06-23 | 2021-08-19 | 강원대학교산학협력단 | Non-destructive Determination Method For Plant Seed Germination And Growth Using Near-Infrared Absorption Spectrum |
Also Published As
| Publication number | Publication date |
|---|---|
| KR20110121838A (en) | 2011-11-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR101206295B1 (en) | Discrimination of geographical origin of grain using near infrared multivariate analysis method | |
| Ambrose et al. | High speed measurement of corn seed viability using hyperspectral imaging | |
| US10073032B2 (en) | Method of simultaneously analyzing amount of nutritional component in various foods having different physicochemical properties and compositions by near-infrared reflectance spectroscopy | |
| Ambrose et al. | Comparative nondestructive measurement of corn seed viability using Fourier transform near-infrared (FT-NIR) and Raman spectroscopy | |
| Wang et al. | Prediction of low heavy metal concentrations in agricultural soils using visible and near-infrared reflectance spectroscopy | |
| Zhao et al. | Near infrared reflectance spectroscopy for determination of the geographical origin of wheat | |
| Chen et al. | Discrimination of Ganoderma lucidum according to geographical origin with near infrared diffuse reflectance spectroscopy and pattern recognition techniques | |
| Feng et al. | X-ray fluorescence application in food, feed, and agricultural science: a critical review | |
| Xu et al. | Determination of rice root density from Vis–NIR spectroscopy by support vector machine regression and spectral variable selection techniques | |
| Shao et al. | Determination of the bruise degree for cherry using Vis-NIR reflection spectroscopy coupled with multivariate analysis | |
| González-Martín et al. | Chilean flour and wheat grain: Tracing their origin using near infrared spectroscopy and chemometrics | |
| Rashmi et al. | Stable isotope ratio analysis in determining the geographical traceability of Indian wheat | |
| Comino et al. | Near-infrared spectroscopy and X-ray fluorescence data fusion for olive leaf analysis and crop nutritional status determination | |
| Xiao et al. | Discrimination of organic and conventional rice by chemometric analysis of NIR spectra: a pilot study | |
| CN107515203A (en) | The research of near infrared technology quantitative analysis rice single grain amylose content | |
| KR100934410B1 (en) | Simple determination of seed weights in crops using near infrared reflectance spectroscopy | |
| Amenta et al. | Traceability of ‘Limone di Siracusa PGI’by a multidisciplinary analytical and chemometric approach | |
| Başayiğit et al. | The prediction of iron contents in orchards using VNIR spectroscopy | |
| CN103822899A (en) | Method for detecting insect fragments in wheat flour based on Terahertz time domain spectra | |
| Lee et al. | Field assessment of macronutrients and nitrogen in apple leaves using a chlorophyll meter | |
| CN106018321A (en) | Establishment method for single corn seed protein detection model and application of establishment method | |
| He et al. | Study on the identification of resistance of rice blast based on near infrared spectroscopy | |
| Nkouaya Mbanjo et al. | Predicting starch content in cassava fresh roots using near-infrared spectroscopy | |
| Nguyen Minh et al. | Authenticity green coffee bean species and geographical origin using near‐infrared spectroscopy combined with chemometrics | |
| Soni et al. | A review of conventional and rapid analytical techniques coupled with multivariate analysis for origin traceability of soybean |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| A201 | Request for examination | ||
| E902 | Notification of reason for refusal | ||
| E701 | Decision to grant or registration of patent right | ||
| GRNT | Written decision to grant | ||
| FPAY | Annual fee payment |
Payment date: 20151012 Year of fee payment: 4 |
|
| FPAY | Annual fee payment |
Payment date: 20160927 Year of fee payment: 5 |
|
| LAPS | Lapse due to unpaid annual fee |