Method for detecting rancidity in oil-containing fruits, seeds and nuts
阅读说明：本技术 用于检测含油果实、种子和坚果中的酸败的方法 (Method for detecting rancidity in oil-containing fruits, seeds and nuts ) 是由 冈瑟·沃尔夫冈·帕尔图夫 埃里希·莱特纳 于 2018-04-20 设计创作，主要内容包括：创建酸败指标表和将酸败指标值分配给含油果实、坚果和种子的吸收或反射光谱(2)包括：-使用光源(3)照射含油果实、坚果或种子(2)的样品,-将反射光和/或透射光投射到光电传感器(4)上,-借助于光电传感器(4)检测吸收或反射光谱,-通过确定挥发性化合物提取样品的成分,-借助于气相色谱法分离样品的挥发性成分,-通过对相关挥发性成分进行质谱检测识别出分离的挥发性成分,-从样品的识别出的挥发性成分确定样品的酸败指标值,-将检测到的样品的吸收或反射光谱分配给酸败指标值,-对代表性数量的样品重复前述步骤,并根据所确定的酸败指标值和分配的吸收或反射光谱形成酸败表。(Creating a rancidity index table and assigning rancidity index values to absorption or reflection spectra of the oleaginous fruits, nuts and seeds (2) comprises: -irradiating the fruit containing oil with a light source (3), -a sample of nuts or seeds (2), -projecting reflected and/or transmitted light onto a photosensor (4), -detecting an absorption or reflection spectrum by means of the photosensor (4), -extracting components of the sample by determining volatile compounds, -separating volatile components of the sample by means of gas chromatography, -identifying separated volatile components by mass spectrometric detection of the relevant volatile components, -determining a rancidity index value for the sample from the identified volatile components of the sample, -assigning the detected absorption or reflection spectrum of the sample to the rancidity index value, -repeating the preceding steps for a representative number of samples and forming a rancidity table from the determined rancidity index value and the assigned absorption or reflection spectrum.)
1. A method for creating a rancidity index table and assigning rancidity index values to absorption or reflection spectra of oily fruits, nuts and seeds (2), comprising the steps of:
-irradiating a sample of the oil bearing fruit, nut or seed (2) with a light source (3),
-projecting the reflected and/or transmitted light onto a photosensor (4),
-detecting an absorption or reflection spectrum in a wavelength range from 900nm to 2500nm, preferably from 900nm to 1700nm, more preferably from 1000nm to 1500nm, by means of the photosensor (4),
-extracting components of the sample by a sample preparation technique based on the determination of volatile compounds from the vapour space above the sample, preferably by solid phase micro-extraction,
-separating the volatile components of the sample by applying gas chromatography techniques,
-identifying the separated volatile components of the sample by mass spectrometric detection of components of the volatile components associated with fat oxidation,
-determining an index value of the rancidity of the sample on the basis of the identified volatile components of the sample,
-assigning at least individual characteristic wavelengths or wavelength ranges of the detected absorption or reflection spectrum of the sample to the index of rancidity,
-repeating the preceding steps for a representative number of samples and forming a rancidity table from the determined rancidity index value and the assigned absorption or reflection spectrum, or from the characteristic wavelength or wavelength range of the detected absorption or reflection spectrum, respectively.
2. The method of claim 1, wherein identifying volatile components of the sample by mass spectrometric detection of components associated with fat oxidation comprises identifying one or more groups of substances/functional groups selected from the group consisting of:
-hydroperoxides
-cyclic hydroperoxides
-saturated, monounsaturated, and diunsaturated aldehydes
Hydrocarbons (alkanes, alkenes)
Alcohols (saturated and unsaturated)
Ketones (saturated and unsaturated)
Short chain fatty acids
-alkylfurans.
3. Method according to claim 1 or 2, characterized in that the identification of the volatile components of the sample by mass spectrometric detection of components related to fat oxidation is performed by creating a chromatogram at a mass-to-charge ratio selected in the range between 20 and 300, preferably at least one mass-to-charge ratio selected among 43, 44, 55, 56, 57, 60, 70, 71, 73, 74, 81, 83, 97.
4. The method of claim 3, wherein the fatty acid chromatogram is created at a mass to charge characteristic of the fatty acid, in particular at a mass to charge ratio of 60.
5. The method of claim 4, wherein a fatty acid index value is determined by integrating over at least a portion of the fatty acid chromatogram.
6. The method according to any one of claims 3 to 5, characterized in that an aldehyde chromatogram is created at the mass to charge ratio characteristic of the aldehyde, in particular at a mass to charge ratio of 44.
7. The method of claim 6, wherein an aldehyde index value is determined by integrating over at least a portion of the aldehyde chromatogram.
8. The method of claim 5 or 7, wherein determining the rancidity index value is performed by forming a sum of the aldehyde index value and the fatty acid index value.
9. The method according to any of the preceding claims, wherein assigning the detected absorption or reflection spectrum of the sample to the rancidity index value is performed by assigning the rancidity index value to at least one of an average, a bandwidth or an individual frequency band of the detected absorption or reflection spectrum.
10. Method according to any of the preceding claims, characterized in that the detection of the absorption or reflection spectrum is achieved by hyperspectral detection by means of the photosensor (4).
11. A method for detecting rancidity in an oil-containing fruit, nut or seed (2), comprising the steps of:
-irradiating the oil-containing fruit, nut or seed (2) with at least one light source (3),
-projecting the reflected and/or transmitted light onto a photosensor (4),
-detecting an absorption or reflection spectrum in a wavelength range from 900nm to 2500nm, preferably from 900nm to 1700nm, more preferably from 1000nm to 1500nm, by means of the photosensor (4),
-providing a rancidity index table according to any of claims 1-10, comprising rancidity index values and assigned absorption or reflection spectra, or characteristic wavelengths or wavelength ranges of said absorption or reflection spectra, respectively,
-assigning the detected absorption or reflection spectrum, or respectively the characteristic wavelength or wavelength range of the detected absorption or reflection spectrum, to the absorption or reflection spectrum of the rancidity indicator table that is most similar to the detected absorption or reflection spectrum, or respectively the characteristic wavelength or wavelength range of the absorption or reflection spectrum,
-determining a value of the rancidity index assigned to the most similar absorption or reflection spectrum or to a characteristic wavelength or wavelength range of the absorption or reflection spectrum, respectively.
12. The method of claim 11, wherein the detected absorption or reflection spectrum is assigned to the most similar absorption or reflection spectrum of the rancidity index table by comparing at least one of an average, a bandwidth, or an individual frequency band of the absorption or reflection spectrum.
13. Method according to any of claims 11 or 12, wherein the oily fruit, nut or seed (2) is separated if the determined rancidity index value exceeds a threshold value.
14. A device (1) for detecting rancid oily fruits, nuts or seeds (2), comprising a light source (3), a photoelectric sensor (4), a computer unit (5) and a classification unit (6), wherein the light source (3) is designed for illuminating the oil-containing fruit, nut or seed (2), the photosensor (4) is connected to the computer unit (3) and is designed for detecting an absorption or reflection spectrum of light reflected from the fruit, nut or seed (2) or transmitted through the fruit, nut or seed (2) and transmitting it to the computer unit (5), and the classification unit (6) is connected to the computer unit (5), the computer unit (5) is designed for controlling the classification unit (6) by performing the method according to any one of claims 11 to 13.
Technical Field
The present invention relates to a method for creating a rancidity index table and assigning rancidity index values to absorption or reflection spectra of oily fruits, nuts and seeds. Furthermore, the present invention relates to a method for detecting rancidity in oil containing fruits, nuts and seeds.
Furthermore, the invention relates to a device for detecting rancid oily fruits, nuts or seeds.
Background
The detection and subsequent sorting of bulk materials by means of photosensors is a widely used method. An embodiment of such a method and such an apparatus for sorting seeds is described, for example, in publication US2013/0278919 a 1. In this known method, the seeds are individually subjected to a spectroscopic examination by illumination with a light source. Subsequently, the absorption or reflection spectrum is captured by the photosensor. Thus, the computer unit analyzes the absorption or reflection spectrum of each seed in the region of interest and calculates the content of a particular component of the seed based on the calibration curve.
For example, the identification of various ones of the individual elements of the bulk material is of interest in order to be able to distinguish between deteriorated elements and non-deteriorated elements of the bulk material. According to the prior art, these methods generally operate in the near infrared range. In order to be able to use those methods in a production facility, the employed photosensors must have a high refresh rate, typically 300Hz or higher. Thus, a high throughput can be ensured while reliably analyzing the composition of each individual element examined. Conventionally, the data collected by the photosensors are analyzed by means of common statistical classification methods such as partial least squares, principal component regression, and the like. This qualitative analysis yields very good results if there is a clear difference between the deteriorated elements and the non-deteriorated elements in the absorption or reflection spectrum.
A disadvantage of this approach is that attempting to separate metamorphic elements from non-metamorphic elements often results in a large number of misclassifications if the absorption or reflection spectra become too similar. A particular disadvantage is that, in the case of subsequent sorting, this leads to a large waste of non-spoiled elements. At the same time, the disadvantage of achieving only a very low detection rate of the deteriorated elements arises. This adverse effect is particularly pronounced in natural products, such as food products, since the natural spectral scattering of the undenatured elements is very broad compared to elements produced in a controlled manner, such as, for example, plastic sheets.
In particular with regard to oily fruits, nuts and seeds which are processed in an automated manner in a production facility, there is considerable interest in automatically distinguishing between rancid and therefore spoiled elements and non-rancid components.
The taste quality of nuts and other seeds or oil-containing fruits is often compromised by fat oxidation, which results in an undesirable rancid taste. Nuts and other oily fruits with a high fat content oxidize fat and as a result can undergo rancidity during storage and processing. This seriously impairs the organoleptic properties (reduced quality of enjoyment) and results in a lower value of the product.
The mechanism of fat oxidation is well known and described in the literature there are two different mechanisms that cause rancidity hydrolysis rancidity is caused by the reaction of water with lipids in the presence of enzymatic activity (lipase). oxidative rancidity can be divided into autoxidation, photooxidation and enzymatic oxidation reactions.among other factors, fatty acid composition is critical in terms of product stability. the stability of unsaturated acids is significantly reduced by increasing the degree of unsaturation. for stearic acid (18:0), oleic acid (18:1, omega-9), linoleic acid (18:2, omega-6) and α -linolenic acid (18:3, omega-3), the rate of oxidation of fatty acids is about 1: 10: 100: 200.
Rancidity may be determined in different ways. Most methods require a large sample in order to use a homogeneous mixture of separately ground oleaginous fruits, nuts or seeds. This involves the disadvantage that valuable information of the individual fruits, nuts or seeds is lost by the respectively homogeneous large amount of oil-containing fruits, nuts or seeds. Another disadvantage of the known methods is that these methods are only laboratory methods and are not suitable for use in automated detection methods and sorting facilities. Furthermore, these laboratory methods are extremely time consuming and, as mentioned above, are suitable for homogenising the product, but not for individual oil-containing fruits, seeds and nuts.
From documents A.Beltr-n, M.Ramos, N.gran re, M.L.Marti n, M.C.Garrigo s, monitoring of oxidation of almond oil by HS-SPME-GC-MS and FTIR, Application of vitamin composition determination to customer evaluation, Food Chemistry, Volume 126,
The study was not performed on individual almonds, but on almonds from multiple almonds, i.e. from homogenized products. There is neither mention nor suggestion of how the results of this study could be used to accomplish the detection and optionally isolation of individual almonds in the product stream. Instead, the study was aimed at proposing a method based on HS-SPME in combination with GC-MS, by means of which it should be possible to rapidly analyze and characterize the volatile components in almond oil resulting from the oxidation of fats. In addition to the measurement by means of HS-SPME/GC-MS, ATR-FTIR spectra of almond oil were also monitored during the oxidation process. It is noted that in HS-SPME-GC-MS analysis, the optimal time for oxidative heat treatment of the sample is seven days, since only significant differences in aldehyde content are discernible after this time, and this time constitutes the "reasonably short duration" of the analysis time. As for additional analysis by ATR-FIR, the measurement of the samples after 1 to 20 days of heat treatment under oxidative conditions was explained, wherein no significant differences in the obtained spectra were still found after one, three and five days of heat treatment time. Only after the fifth day of heat treatment, a spectral change was observed, indicating a gradual oxidation of the sample. Measurements using HS-SPME/GC-MS or using ATR-FTIR, respectively, are performed independently of each other and the measurement results from both types of measurements are only used to determine whether the results of the respective other type of measurement seem reasonable. However, the results of the two types of measurements are not linked. In particular, no rancidity indicator or any other indicator is created from the measurement results.
In the literature Borr-s E, Ferr J, Boqu R, strains M, Acenia L, Calvo A, Butoo.prediction of organic oil sensory descriptors using the atomic data fusion and Partial Least Squares (PLS) regression. Talanta.2016Aug 1; 155:116-23.Epub2016, 20/4, mass spectrometry (HS-MS), fourier transform mid-infrared spectroscopy (FT-MIR), and UV-visible spectrophotometry (UV-vis) were used to predict olive oil taste parameters. A multivariate calibration model was created based on measurements from 343 olive oil samples from four consecutive harvests using partial least squares regression. The HS-MS and FT-MIR results are either evaluated separately or, to improve the prediction model, are linked by "data fusion" which is the combination of two resulting matrices for two different types of measurements to form one matrix. However, during this examination, it becomes apparent that "data fusion" also does not provide a useful prediction of rancidity in olive oil, since only 10% (|) samples with rancid olive oil can be correctly detected. Due to the effect of this undesirable result, the authors of D2 indicated that it was not possible to create a usable rancidity model from the measurement data.
It is therefore an object of the present invention to provide a method which avoids the disadvantages of the prior art and which enables automatic detection of rancidity in individual oily fruits, nuts and seeds.
Disclosure of Invention
According to the invention, this object is achieved by a method for creating a rancidity index table and assigning rancidity index values to absorption or reflection spectra of oily fruits, nuts and seeds, comprising the steps of:
-illuminating a sample of an oil bearing fruit, nut or seed with a light source,
projecting the reflected and/or transmitted light onto a photosensor,
-detecting the absorption or reflection spectrum of the sample in the wavelength range from 900 to 2500nm, preferably from 900 to 1700nm, more preferably from 1000 to 1500nm, by means of a photosensor,
extracting components of the sample by a sample preparation technique based on the determination of volatile compounds from the vapour space above the sample, preferably by solid phase microextraction,
-separating the volatile components of the sample by applying gas chromatography techniques,
identifying the volatile components separated in the sample by mass spectrometric detection of the components of the volatile components associated with fat oxidation,
-determining a sample's index of rancidity from the identified volatile components of the sample,
assigning at least individual characteristic wavelengths or wavelength ranges of the detected absorption or reflection spectrum of the sample to the spoilage indicator value,
-repeating the preceding steps for a representative number of samples and forming a rancidity table from the determined rancidity index value and the assigned absorption or reflection spectrum, or from the characteristic wavelength or wavelength range of the detected absorption or reflection spectrum, respectively.
Compared to analytical, very time consuming methods in the laboratory, such as the known methods described above, the method according to the invention is suitable for automatic detection and sorting depending on the level of rancidity of oil-containing fruits, seeds and nuts, wherein a high production yield of individual fruits is detectable and the detected rancid fruits are individually separable from the product stream in a sorting facility.
In a preferred embodiment of the invention, identifying volatile components of the sample by mass spectrometric detection of components associated with fat oxidation comprises identifying one or more groups of substances/functional groups selected from the group consisting of:
-hydroperoxides
-cyclic hydroperoxides
-saturated, monounsaturated, and diunsaturated aldehydes
Hydrocarbons (alkanes, alkenes)
Alcohols (saturated and unsaturated)
Ketones (saturated and unsaturated)
Short chain fatty acids
-alkylfurans.
Identification of volatile components of the sample by mass spectrometric detection of components associated with fat oxidation may in particular be carried out by creating a chromatogram at a mass to charge ratio selected in the range between 20 and 300, preferably at least one mass to charge ratio selected in 43, 44, 55, 56, 57, 60, 70, 71, 73, 74, 81, 83, 97.
In a particular embodiment of the invention, identifying volatile components of the sample by mass spectrometric detection of components associated with fat oxidation includes creating a fatty acid chromatogram at a first mass-to-charge ratio and an aldehyde chromatogram at a second mass-to-charge ratio.
The samples used according to the method of the invention are chosen in terms of quantity and characteristics in such a way that a representative cross section of possible rancidity can be determined. This can be ensured in particular by appropriate sample selection and preparation (e.g. storage at elevated temperature for different time periods).
In order to measure the absorption or reflection spectra of the oily fruits, nuts or seeds, respectively, by hyperspectral imaging (HSI) for the determination of rancidity and the subsequent preferred classification of the invention, detailed information about the individual oily fruits, nuts or seeds, respectively, is necessary for a suitable calibration model. This is ensured by the method according to the invention, whereby the separation of the rancid, oily fruit, nut or seed, respectively, can be achieved more than ever. In particular, the present invention provides a highly accurate and unique reference analysis by means of which the actual oxidation state of individual nuts/oily fruits/seeds can be determined and extreme values thereof can be collected.
One embodiment of the method according to the invention is based on the accumulation of volatile compounds in the vapor space above the sample, in particular on headspace solid phase microextraction (HS-SPME), which is coupled with gas chromatography with mass selective detection. This combination of analytical methods is abbreviated HS-SPME-GC-MS. However, the invention is not limited to this embodiment, but the method according to the invention basically comprises measuring and determining volatile compounds in the vapor space of the sample, separating the individual components of the volatile compounds by gas chromatography, identifying at least individual ones of the separated volatile compounds by mass selective detection, and selectively using the identified volatile compounds to determine a rancidity indicator.
Fatty oxidation of unsaturated fatty acids begins with the formation of hydroperoxides and produces a large number of different chemical structures and functional groups. Thus, these groups of substances can potentially be used as marker compounds for measuring the degree of fat oxidation.
For hyperspectral analysis it is recommended to consider all wavelengths in the near infrared range that are able to measure the structural properties as described above.
Compounds which cause an undesired rancid taste are, for example, aldehydes resulting from the cleavage of fatty acid chains after the formation of hydroperoxides. As oxidation proceeds, aldehydes may form free fatty acids, which may also contribute to further undesirable organoleptic properties of the rancid nuts. To assess the quality of individual oil-containing fruits or seeds, these oil-containing fruits or seeds (with different rancidity and different origins) were first measured by HSI, individually labeled in time, and packaged and analyzed by HS-SPME-GC-MS as follows:
individual oil-containing fruits, nuts or seeds are ground and a suitable representative amount of sample, for example 300mg, is weighed into a glass container of suitable size and sealed in an airtight manner. A glass coated magnetic stirrer may be included in the glass container. The accumulation of volatile components is achieved by suitable techniques based on headspace analysis of suitable adsorbing or absorbing materials capable of reversibly binding volatile organic compounds. Furthermore, desorption takes place thermally at elevated temperatures, preferably directly in the sample injection system of the gas chromatography system. Separation of volatile compounds occurs on a high resolution capillary with a suitable stationary phase and temperature program capable of separating analytes. Detection occurs via mass selective detection, such that mass spectra can be captured across the entire mass range of the relevant target compound to enable unambiguous identification of the compound. The mass spectrum is detected in a scanning mode, and the scanning range of the mass-to-charge ratio (m/z) is preferably 20-300.
Further information can be obtained by extracting the selected mass-to-charge ratio. It has been shown that m/z 44 constitutes a universal and selective fragment for linear and saturated aldehydes, which is well suited for rancidity determination. Additional information about potential rancidity may be obtained via free fatty acids using a mass to charge ratio m/z of 60, according to one embodiment of the method an aldehyde chromatogram is thus created at m/z of 44, and a fatty acid chromatogram is created at m/z of 60.
In one embodiment of the invention, the rancidity index value may be calculated by integrating the peak of an aldehyde chromatogram having an m/z of 44 and the peak of a fatty acid chromatogram having an m/z of 60. The value obtained may be expressed as an aldehyde index value or a fatty acid index value or a total rancidity index. For ease of reading, the sum of the peak ranges is divided by a fixed number to obtain a more manageable number. By analyzing a large number of different samples of different origin and quality, a wide range of values of rancidity index for calibrating the model can be determined and verified separately.
Thus, the method according to the invention uses a quantitative method which does not involve finding distinct distinguishing features in the absorption or reflection spectrum, but rather creates a correlation between a small but still significant difference in the absorption or reflection spectrum and a reference from a laboratory. Therefore, absorption or reflection spectra are not used to classify oil-bearing fruits and seeds into two categories, good (products with little rancidity reaction) and bad (products with rancidity reaction), but create an index of rancidity. As a result, an advantage is obtained that the degree of rancidity can be quantitatively detected accordingly. It is particularly advantageous that only in subsequent steps one or more threshold values for the rancidity index may be provided, from which threshold values the oily fruit, nut or seed, respectively, is classified as no longer meeting the quality criterion or as falling into a different quality level, respectively. In particular, this allows the advantage of very simple adaptation to different quality requirements for oil-containing fruits and seeds.
The method according to the invention also provides the following advantages: the first time, the identifiability of rancidity in oil-bearing fruits and seeds by means of spectroscopic methods can be demonstrated. For this reason, chemical markers directly related to rancidity were found in the laboratory. Thus, using a statistical correlation method, the absorption or reflection spectrum of the photosensor that has been recorded for the selected sample is again correlated with the rancidity index determined in the laboratory for the same sample, so that the rancidity index can be calculated directly on the basis of the absorption or reflection spectrum.
Drawings
Advantageous embodiments as well as alternative embodiment variants of the method according to the invention will be explained in more detail below with reference to the drawings.
FIG. 1 shows chromatograms of a rancid sample and a fresh sample.
FIG. 2 shows chromatograms of extracted m/z 44 (saturated aldehydes) for rancid and fresh nuts.
FIG. 3 shows chromatograms of extracted m/z 60 (fatty acid) for rancid and fresh nuts.
Fig. 4 shows schematically an apparatus for detecting rancid oily fruits, nuts or seeds.
Detailed Description
The method according to the invention provides for assigning rancidity index values to individual oily fruits, nuts and
In order to determine the index of the rancidity of the fruit, nut or
In a preferred embodiment variant, the aldehyde index value is determined by integration over at least a part of the determined aldehyde chromatogram and the fatty acid index value is determined by integration over at least a part of the determined fatty acid chromatogram.
For illustrative purposes, FIG. 1 shows chromatograms of a rancid sample and a fresh sample analyzed according to the previously described method, with a mass spectrometer operated for detection in a scan mode of mass-to-charge ratio (m/z) over a relevant mass range of, for example, 20-300.
For illustrative purposes, FIG. 2 shows an example of an aldehyde chromatogram for a rancid sample at a first mass-to-charge ratio (m/z) of 44 compared to a fresh sample. The determination of the aldehyde index value as described above apparently results in a significantly higher aldehyde index value for the aldehyde chromatogram of the rancid sample than for the good sample.
For illustrative purposes, FIG. 3 shows an example of a fatty acid chromatogram of a rancid sample at a second mass to charge ratio (m/z) of 60 compared to a fresh sample. Consistent with the aldehyde index values set forth in fig. 2, the determination of fatty acid index values as described above clearly yields significantly higher fatty acid index values for the fatty acid chromatogram of the rancid sample than for the fresh sample.
These steps are repeated for representative various oleaginous fruits, nuts or seeds and a rancidity index table is created from the determined rancidity index values and the assigned absorption and reflection spectra.
In a preferred embodiment variant of the method, assigning the absorbance or reflection spectrum of the detected oleaginous fruit, nut or seed to the rancidity index value is achieved by assigning the rancidity index value to at least one of an average, a bandwidth or a separate frequency band of the detected absorbance or reflection spectrum. In doing so, certain ranges or averages of the corresponding absorption or reflection spectra are defined as ranges characterizing the rancidity of the oil-containing fruit, nut or
Furthermore, the present invention provides a method for detecting the rancidity of an oleaginous fruit, nut or
The light reflected from the fruit, nut or seed or transmitted through the fruit, nut or seed, respectively, is then projected onto a photosensor, which detects an absorption or reflection spectrum in the near infrared range, preferably from 900 to 1700 nm. In a particularly preferred embodiment variant, the absorption or reflection spectrum is detected by a photosensor in the range from 1000 to 1500 nm. Preferably, the detection of the absorption or reflection spectrum is effected by hyperspectral detection by means of a hyperspectral camera.
In a further step, the method utilizes a rancidity table of the aforementioned method comprising rancidity index values and assigned absorption or reflection spectra or characteristic ranges and/or wavelengths of those spectra, respectively. The absorption or reflection spectrum detected in the method is then compared with the absorption or reflection spectrum included in the rancidity indicator table. In this way, it is achieved that the detected absorption or reflection spectrum is assigned to the absorption or reflection spectrum of the rancidity index table which is most similar to the detected absorption or reflection spectrum. This allows to determine the index of rancidity assigned to the most similar absorption or reflection spectrum.
This method offers the advantage that by means of the previously described method for creating a rancidity index table according to the invention, it is possible to detect rancidity of oily fruits, nuts or seeds in a production facility by means of a previous calibration, wherein the rancidity of individual fruits is detected instead of only a uniform product flow.
According to a preferred embodiment variant of the detection method according to the invention, the assignment of the detected absorption or reflection spectrum to the absorption or reflection spectrum of the rancidity indicator table that is most similar to the detected absorption or reflection spectrum is achieved by comparing at least one of the average, the bandwidth or the individual frequency bands of the absorption or reflection spectrum. In this case, certain ranges or averages of the corresponding absorption or reflection spectra are defined as ranges characterizing the rancidity of the oil-containing fruit, nut or seed and are used for the comparison. This advantageously increases the accuracy of the processing.
Furthermore, the detection method provides the advantage that in a further method stage at least one threshold value can be determined and exceeding the at least one threshold value results in the oily fruit, nut or seed being discarded or sorted in a distinctive manner. This provides the advantage that the definition of the threshold value makes it possible to adapt to different quality requirements of the oil-containing fruit, nut or seed, respectively.
Finally, the invention also comprises a device for detecting rancid oily fruits, nuts or seeds, as will be described in detail below, which is designed to carry out a detection method according to the invention for creating a rancid index table, using a rancid index table created according to the method of the invention.
Fig. 4 shows in a schematic view a device 1 for detecting a rancid oily fruit, nut or
The photosensor 4 is connected to the
The
Downstream of the photoelectric sensor 4, the fruit, nut or
In summary, the invention as defined in the independent claims and in the preferred embodiments is based on the following concepts and advantages:
the detection and classification of bulk materials by means of photosensors using hyperspectral imaging (HSI) is a widely used method. In doing so, the sample is illuminated with broadband light, and the reflected light is spectrally detected and examined by a photosensor, preferably in the near infrared range. Based on the evaluation of the spectra (amplitude, frequency), characteristics and components are inferred, respectively. The identification of these properties, components, forms the basis for distinguishing good and bad products in the classification process. The identification of these characteristics/components is based on a qualitative, relative approach. In the modeling process, the photometrically detected spectral profile of a good product is compared to the spectral profile of a bad product. In doing so, the spectral regions are sought as follows: in this region, the difference in characteristics is very large or the correlation with the substance sought is very high compared to other substances. The identified regions are then selected for classification processing and the corresponding spectra are normalized. That is, the absolute amplitude of the spectrum is eliminated, and only the difference between the spectra of good and bad products is used in the classification for the determination.
Due to this process, misclassification can occur for the following reasons:
if the difference between the two spectra is very small, the small difference is enhanced in the normalization process to the extent that the signal-to-noise ratio will increase drastically, which greatly increases the uncertainty of the decision.
The difference in spectra is determined based on a reference sample of good/bad samples. However, for natural products with spectral scattering, this comparison suffers from a high level of uncertainty and may vary from batch to batch.
The result of such misclassification is an insufficient separation of good and spoiled food.
This problem is particularly present when oil-containing fruits, nuts and seeds are automatically processed in production facilities. In such automated processes, the emphasis is on the desire to automatically distinguish rancid, and therefore spoiled, elements from non-rancid elements. Individual low quality fruits/nuts/seeds will be removed in an automated manner at high production rates even before the product is further processed (pressing, grinding, peeling, etc.).
The proposed solution is based on a quantitative method which eliminates the drawbacks of the previous methods, since the classification information is not, as in the past, based on a spectral comparison of the rancid oily fruit/nut/seed with a comparative amount of non-rancid, but rather the chemical initiators of rancidity (fat oxidation, hydrolysis) and the substances (aldehydes, etc.) resulting therefrom are examined from the point of view of spectral fingerprints.
Since rancidity is not a binary quantity, i.e. it is not sufficient to merely distinguish between rancidity/non-rancidity, but rancidity is present to a different extent, a rancidity index table developed according to the invention is used which assigns amplitudes in the respective spectral ranges to, for example, a rancidity value of 0-100%.
The rancidity index table is developed based on a large number of statistical oil bearing fruits/nuts/seeds using available analytical laboratory methods (e.g. gas chromatography etc.).
According to the invention, the rancidity index is now used by using a photoelectric sensor, preferably a hyperspectral camera, of the sorting facility for detecting rancidity in oily fruits, nuts and seeds, wherein the photoelectric sensor may be calibrated using a rancidity index table of the sorting facility. That is, based on the table, the quantitative degree of rancidity (e.g., 0-100%) can be automatically inferred from absolute amplitudes in the spectrum in the appropriate spectral range.
The characteristics of the detection method based on the rancidity index table created according to the invention and of the classification facilities performing the detection method separately are characterized by the following advantages:
high processing speed and certainty of decision making in the online classification process due to the evaluation of the rancidity index table;
-based on the current state of offline laboratory techniques, a high quality rancidity indicator table can be used online in the classification process;
the classification of rancidity is not a binary quantity (rancidity/non-rancidity), but a similar quantity that can be traced back based on the amplitude at a certain wavelength or respectively on the average of the amplitudes over a range of wavelengths;
based on the rancidity value, it is possible to classify according to different qualities (x% rancidity) and thus to enable the use of different quality levels;
-accurately setting a classification limit in the classification facility.
- 上一篇：一种医用注射器针头装配设备
- 下一篇：分析和治疗方法和组合物及其用途