Accurate determination method and system for residual quantity of fruit and vegetable bactericide
By analyzing the Raman spectral data of fruits and vegetables at different depths, the degree of matrix interference of fungicides inside fruits and vegetables is obtained and corrected, which solves the problem of inaccurate determination of fungicide residues in fruits and vegetables and achieves high-precision determination of fungicide residues.
Patent Information
- Application Number
- CN202510982394.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-16
- Publication Date
- 2025-09-16
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the existing technology, the determination of fungicide residues in fruits and vegetables has accuracy problems, mainly because the complexity of the fruit and vegetable matrix components causes deviations in Raman characteristic signals, affecting the accurate acquisition of fungicide residues.
By obtaining Raman spectral data of fruits and vegetables at different depths, the Raman signal intensity change trend of each characteristic peak and the degree of matrix interference were analyzed. The initial fungicide residue was corrected based on the degree of matrix interference to obtain the final fungicide residue.
It achieves high-precision determination of fungicide residues, eliminates the interference of fruit and vegetable matrices on Raman signals, improves measurement accuracy, and avoids underestimation or overestimation caused by single-depth measurement.
Smart Images
Figure CN120651802A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of spectral testing technology, and in particular to a method and system for accurately determining the residual amount of a fungicide in fruits and vegetables. Background Art
[0002] During the transportation of fruits and vegetables, fungicides are often used to extend their shelf life. Excessive or improper use of fungicides can easily cause them to remain on the surface or inside the fruits and vegetables. In the prior art, Raman spectroscopy is usually used to determine the amount of fungicide residues in fruits and vegetables. However, due to the complex composition of the fruit and vegetable matrix, the tissue structure is rich in water, polysaccharides, organic acids, flavonoids, pigments and other natural products, which may undergo physical adsorption or chemical reactions with fungicides, resulting in deviations in the Raman characteristic signal, such as enhancement (such as surface enhancement effect) or attenuation (such as fluorescence flooding, energy transfer, etc.), thereby affecting the accurate acquisition of the fungicide residue amount of the fungicide characteristic peak, thereby reducing the accuracy of the fruit and vegetable fungicide residue determination. Summary of the Invention
[0003] In order to solve the technical problem that the existing fungicide residual amount of the fungicide characteristic peak is inaccurate, thereby affecting the accuracy of the fungicide residual amount measurement, the purpose of the present invention is to provide a method and system for accurately measuring the residual amount of fruit and vegetable fungicides. The technical solutions adopted are as follows:
[0004] In a first aspect of the present invention, a method for accurately determining the residual amount of a fungicide in fruits and vegetables is provided, comprising:
[0005] Obtain Raman spectral data of fruits and vegetables at different depths;
[0006] According to the characteristic changes of the Raman spectrum data of each characteristic peak at each depth, the Raman signal intensity change trend of each characteristic peak is obtained;
[0007] The degree of influence of matrix interference on each depth of each characteristic peak is obtained based on the difference in trend change differences corresponding to all depths before and after a certain depth of each characteristic peak; the trend change difference represents the difference between the Raman signal intensity change trend of the characteristic peak and the standard Raman signal intensity change trend;
[0008] Correcting the initial fungicide residue measured by the Raman spectrum data at each depth of each characteristic peak according to the degree of matrix interference and the characteristic changes at each depth of each characteristic peak;
[0009] According to the corrected fungicide residue at each depth of each characteristic peak and the degree of matrix interference, the final fungicide residue of each characteristic peak is obtained.
[0010] In an exemplary embodiment, the process of acquiring the feature change includes:
[0011] Determining the characteristic value of each characteristic peak at each depth from the Raman spectrum data of each characteristic peak at each depth;
[0012] The characteristic variation of each characteristic peak at each depth is obtained from the difference in characteristic values of two adjacent depths of each characteristic peak.
[0013] In an exemplary embodiment, the process of obtaining the characteristic value includes:
[0014] Obtaining a peak shape characteristic according to the peak area and peak width of a target characteristic peak at a target depth; the target characteristic peak is any characteristic peak, and the target depth is any depth; the peak shape characteristic is proportional to the peak area and inversely proportional to the peak width;
[0015] Obtaining a Raman frequency shift difference between each signal point and a signal point corresponding to the peak at a target depth of the target characteristic peak, and combining the Raman intensity of each signal point at the target depth of the target characteristic peak to obtain a peak distribution concentration feature at the target depth of the target characteristic peak; the peak distribution concentration feature is inversely proportional to the Raman frequency shift difference and directly proportional to the Raman intensity;
[0016] The peak shape feature and the peak distribution concentration feature are integrated to obtain a characteristic value of a target depth of a target characteristic peak.
[0017] In an exemplary embodiment, the process of acquiring the Raman signal intensity variation trend includes:
[0018] Obtaining the degree of fluctuation of the characteristic change at each depth of each characteristic peak;
[0019] According to the fluctuation degree, a Raman signal intensity variation trend of each characteristic peak is obtained, and the Raman signal intensity variation trend is inversely proportional to the fluctuation degree.
[0020] In an exemplary embodiment, the process of acquiring the standard Raman signal intensity variation trend includes:
[0021] Obtaining Raman spectral data samples of a plurality of standard samples; the standard samples are fruits and vegetables whose matrices do not interfere with the characteristic peaks of the fungicide;
[0022] According to the characteristic changes of the Raman spectrum data samples at each depth of each characteristic peak of each standard sample, the Raman signal intensity change trend of each characteristic peak of each standard sample is obtained;
[0023] The average of the Raman signal intensity variation trends of the same characteristic peak of all standard samples is calculated as the standard Raman signal intensity variation trend of the corresponding characteristic peak.
[0024] In an exemplary embodiment, the process of obtaining the degree of matrix interference includes:
[0025] Obtaining a first trend change difference corresponding to all depths before a target depth of a target characteristic peak, and a second trend change difference corresponding to all depths after the target depth of the target characteristic peak; the target characteristic peak is any characteristic peak, and the target depth is any depth;
[0026] The relative deviation of the second trend change difference with respect to the first trend change difference is obtained to obtain the degree of matrix interference influence on the target depth of the target characteristic peak.
[0027] In an exemplary embodiment, the process of correcting the initial fungicide residue measured by the Raman spectrum data at each depth of each characteristic peak includes:
[0028] The target depth correction coefficient of the target characteristic peak is obtained by fusing the matrix interference influence degree and the characteristic change difference of the target characteristic peak; the correction coefficient is proportional to the matrix interference influence degree and the characteristic change difference; the target characteristic peak is any characteristic peak, and the target depth is any depth; the characteristic change difference is the difference between the characteristic change of the target depth of the target characteristic peak and the standard characteristic change;
[0029] The initial fungicide residue corresponding to the target depth of the target characteristic peak is corrected according to the correction coefficient.
[0030] In an exemplary embodiment, the process of acquiring the standard feature change includes:
[0031] A standard characteristic peak is obtained, where the standard characteristic peak is a characteristic peak corresponding to the standard Raman signal intensity change trend; and the standard characteristic change is a characteristic change of a target depth of the standard characteristic peak.
[0032] In an exemplary embodiment, the process of obtaining the terminal sterilant residue includes:
[0033] Taking the matrix interference influence degree sequence of the target characteristic peak as the weight, the weighted average value of the fungicide residue sequence of the target characteristic peak is obtained as the final fungicide residue of the target characteristic peak; the target characteristic peak is any characteristic peak, the matrix interference influence degree sequence is composed of the matrix interference influence degrees of each depth of the target characteristic peak, and the fungicide residue sequence is composed of the corrected fungicide residues at each depth of the target characteristic peak.
[0034] In a second aspect of the present invention, a system for accurately measuring the residual amount of a fruit and vegetable fungicide is provided, comprising: a memory and a processor; the memory is connected to the processor; the memory is used to store program instructions; and the processor is used to implement the above-mentioned method for accurately measuring the residual amount of the fruit and vegetable fungicide when the program instructions are executed.
[0035] The present invention has the following beneficial effects: by acquiring Raman spectral data at different depths of fruits and vegetables, the present invention obtains the degree of interference and influence of the fruit and vegetable matrix on the fungicide at different depths inside the fruits and vegetables. Based on the degree of matrix interference and the characteristic changes of the Raman spectral data at each depth, the initial fungicide residue at each depth is corrected to obtain the final fungicide residue at the characteristic peak. This effectively distinguishes and weakens the interference of the fruit and vegetable matrix components on the Raman signal, eliminates the influence of the fruit and vegetable matrix on the fungicide Raman signal, and improves the accuracy of obtaining the fungicide residue at the fungicide characteristic peak, thereby achieving high-precision determination of the true fungicide residue. At the same time, acquiring Raman spectral data at multiple different depths of fruits and vegetables can avoid underestimation or overestimation of the fungicide residue caused by single-depth measurement, thereby achieving accurate determination of the overall residue under spatial integration. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] Figure 1 This is a flow chart of a method for accurately determining the residual amount of a fungicide in fruits and vegetables provided by one embodiment of the present invention;
[0037] Figure 2 This is a Raman spectrum of fruits and vegetables at a certain depth provided by an embodiment of the present invention;
[0038] Figure 3 is a flow chart of obtaining feature changes provided by one embodiment of the present invention;
[0039] Figure 4 is a flow chart for obtaining characteristic values provided by one embodiment of the present invention;
[0040] Figure 5 This is a flowchart for obtaining a Raman signal intensity variation trend according to an embodiment of the present invention;
[0041] Figure 6 This is a flow chart for obtaining a standard Raman signal intensity variation trend provided by one embodiment of the present invention;
[0042] Figure 7 This is a flow chart for obtaining the degree of matrix interference provided by one embodiment of the present invention;
[0043] Figure 8 This is a diagram of the correction process of the fungicide residual amount provided by one embodiment of the present invention. DETAILED DESCRIPTION
[0044] To further illustrate the technical means and effects employed by the present invention to achieve its intended objectives, the following detailed description of the specific embodiments, structures, features, and effects of the present invention is provided in conjunction with the accompanying drawings and preferred embodiments. In the following description, references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
[0045] Unless otherwise defined, all technical and scientific terms used herein have the same meanings as commonly understood by those skilled in the art to which this invention pertains. The data and information collected in this application were obtained with full consent and authorization.
[0046] This embodiment provides a method for accurately measuring the residual amount of fungicide on fruits and vegetables, used to detect the residual amount of fungicide sprayed on the surface of fruits and vegetables. Because fruits and vegetables are often sprayed with only one fungicide, such as prochloraz, a common fungicide used during transportation to extend the shelf life of citrus fruits, this embodiment is applicable to scenarios where only one fungicide is sprayed on the surface of fruits and vegetables, and this embodiment measures the residual amount of that fungicide on the surface of fruits and vegetables.
[0047] It should be understood that the number of characteristic peaks of different fungicides may be different. Some fungicides have only one characteristic peak in the Raman spectrum, while some fungicides have multiple characteristic peaks. For example, Prochloraz has a peak at 620 cm in its Raman spectrum. -1 (benzene ring), 1005cm -1 (CN), 1170cm -1 (NC=O), 1290cm -1 (C-Cl), 1600 cm -1 (benzene ring) and other places have characteristic peaks; Azoxystrobin, in its Raman spectrum, at 1000cm -1 (COC), 1180cm -1 (CN), 1600cm -1 (benzene ring) and other places have characteristic peaks.
[0048] like Figure 1 As shown, this embodiment provides a method for accurately determining the residual amount of a fungicide in fruits and vegetables, comprising the following steps:
[0049] Step S1: Acquire Raman spectrum data of fruits and vegetables at different depths;
[0050] Step S2: obtaining a Raman signal intensity variation trend of each characteristic peak based on characteristic variations of the Raman spectrum data at each depth of each characteristic peak;
[0051] Step S3: Obtaining the degree of matrix interference influence on each depth of each characteristic peak based on the difference in trend change corresponding to all depths before and after a certain depth of each characteristic peak;
[0052] Step S4: correcting the initial fungicide residue measured by the Raman spectrum data at each depth of each characteristic peak according to the degree of matrix interference and the characteristic changes at each depth of each characteristic peak;
[0053] Step S5: The final fungicide residue of each characteristic peak is obtained based on the corrected fungicide residue at each depth of each characteristic peak and the degree of matrix interference.
[0054] Each step is described in detail below with reference to the accompanying drawings.
[0055] Step S1: Acquire Raman spectrum data of fruits and vegetables at different depths.
[0056] To accurately measure fungicide residues in fruits and vegetables, a confocal Raman spectroscopy system with adjustable focus is used to collect multi-layer depth spectra. The confocal optical system forms a tiny laser spot within the fruit and vegetable sample. By precisely controlling the focal depth along the Z axis, Raman scattering signals, or Raman spectral data, are collected at different depths within the fruit and vegetable.
[0057] Before measurement, the fruit and vegetable samples to be tested can undergo simple surface pretreatment, such as washing and natural drying, to remove free moisture and surface dust, which can interfere with the Raman signal. The fruit and vegetable samples are then sliced and the cut surface is fixed flatly on the Raman system's stage to ensure a stable sample position and no displacement during the laser focusing and Z-axis scanning process.
[0058] During measurement: A high numerical aperture microscope objective is installed on the confocal Raman microscope to improve the spatial resolution and Raman signal acquisition efficiency. The starting focus position (the surface of the fruit and vegetable sample) is set, and the Z axis is scanned layer by layer in the vertical direction with a step size of 1-10μm (various depths such as: 0μm, 5μm, 10μm, 15μm...), and the Raman spectrum at each depth position is collected in sequence. At each focal depth, an integration time of (0.5-5s) is set to ensure that the characteristic peak signal of the fungicide is captured.
[0059] After measurement: The collected spectral data at different depths are integrated and visualized to generate Raman spectra at each depth. Figure 2 As shown, it is a Raman spectrum of fruits and vegetables at a certain depth. The horizontal axis of the Raman spectrum of fruits and vegetables is Raman frequency shift (also called Raman displacement or wave number), and the vertical axis is Raman intensity (also called Raman scattering intensity).
[0060] Step S2: Obtaining a Raman signal intensity variation trend of each characteristic peak based on characteristic variations of the Raman spectrum data at each depth of each characteristic peak.
[0061] Fruit and vegetable matrix components differ significantly from fungicides in their chemical structures, resulting in distinct peak positions, intensities, and shapes in Raman spectra. Matrix components of fruits and vegetables primarily include water, polysaccharides, organic acids, pigments, and cellulose. Raman signals are often concentrated in specific wavelength bands and are often broad or form a background. Fungicide molecules, on the other hand, typically possess specific functional group structures, resulting in relatively clear, well-positioned, and reproducibly identifiable peaks in Raman spectra. To achieve qualitative identification and quantitative analysis, the fungicide's Raman signature peaks must first be identified and extracted.
[0062] In one exemplary embodiment, a standard Raman spectrum of the fungicide is obtained. This standard Raman spectrum is not affected by components of the fruit or vegetable matrix, and the wavenumber range of the characteristic peaks is determined from this spectrum. The positions of the characteristic peaks of the fungicide are then determined from the Raman spectrum data obtained at different depths of the fruit or vegetable, thereby obtaining the characteristic peaks of the fungicide.
[0063] The area distribution of each characteristic peak of the fungicide is concentrated, the peak intensity is high, and the morphology is clear, while the Raman signals of the natural components of fruits and vegetables are mostly weak in intensity, blunt in spectrum, and discrete in distribution. Therefore, the characteristic changes of the Raman spectrum data at each depth of each characteristic peak of the fungicide are obtained. In an exemplary embodiment, Figure 3 As shown, a specific acquisition process of feature changes is given as follows:
[0064] Step S2 - 1 : Determine the characteristic value of each characteristic peak at each depth from the Raman spectrum data of each characteristic peak at each depth.
[0065] According to the relevant characteristics of the Raman spectrum data of each depth of each characteristic peak, the characteristic value of each depth of each characteristic peak is obtained. Figure 4 As shown, a specific process of obtaining the eigenvalue is given as follows:
[0066] Step S2-1-1: Obtain peak shape characteristics based on the peak area and peak width at the target depth of the target characteristic peak.
[0067] For ease of explanation, any characteristic peak of the fungicide is set as a target characteristic peak, and any depth is set as a target depth.
[0068] Obtain the peak area of the target characteristic peak at the target depth, where the peak area is essentially the area under the curve below the target characteristic peak and above the horizontal coordinate as the baseline. The peak area is the integral value of the curve segment corresponding to the target characteristic peak relative to the horizontal coordinate.
[0069] Obtain each signal point on the horizontal axis of the target characteristic peak. Each signal point includes the starting point and ending point of the target characteristic peak on the horizontal axis, as well as each signal point between the starting point and the ending point. Obtain the Raman intensity of each signal point on the horizontal axis of the target characteristic peak, integrate these discrete Raman intensities, and the obtained integral result is the peak area of the target characteristic peak. Obtain the peak width of the target characteristic peak at the target depth. The peak width is essentially the number of signal points on the horizontal axis of the target characteristic peak at the target depth.
[0070] The peak shape characteristics of the target characteristic peak are directly proportional to the peak area and inversely proportional to the peak width. The larger the peak area and the smaller the peak width, the more the target characteristic peak has the characteristics of a wave crest, that is, the more distinct the peak shape characteristics of the target characteristic peak. In an exemplary embodiment, a specific method for quantifying the peak shape characteristics of the target characteristic peak is as follows: the ratio of the peak area to the peak width of the target characteristic peak is calculated, and the result obtained is the peak shape characteristics of the target characteristic peak.
[0071] Step S2-1-2: Obtain the Raman frequency shift difference between each signal point in the target depth of the target characteristic peak and the signal point corresponding to the peak, and combine the Raman intensity of each signal point in the target depth of the target characteristic peak to obtain the peak distribution concentration characteristics of the target depth of the target characteristic peak.
[0072] Determine the position of the signal point on the horizontal axis corresponding to the peak value of the target characteristic peak, where the peak value is the maximum Raman intensity in the target characteristic peak. Obtain the Raman frequency shift difference between each signal point in the target depth of the target characteristic peak and the signal point corresponding to the peak value. In essence, it is the distance on the horizontal axis between each signal point of the target characteristic peak in the target depth and the signal point corresponding to the peak value. Combined with the Raman intensity of each signal point of the target characteristic peak in the target depth, the peak distribution concentration feature of the target depth of the target characteristic peak is obtained. The higher the Raman intensity of each signal point of the target characteristic peak in the target depth, the smaller the Raman frequency shift difference, indicating that the distribution of the target characteristic peak at the target depth is more concentrated, and the peak distribution concentration feature is more obvious. Therefore, the peak distribution concentration feature is inversely proportional to the Raman frequency shift difference and directly proportional to the Raman intensity. It should be understood that the Raman frequency shift difference between the signal point corresponding to the peak and itself is 0, and does not participate in the calculation of the peak distribution concentration feature.
[0073] Step S2-1-3: Fuse the peak shape feature and the peak distribution concentration feature to obtain the characteristic value of the target depth of the target characteristic peak.
[0074] The target depth characteristics of the target characteristic peak are reflected from two aspects: the peak shape characteristics and the peak distribution concentration characteristics of the target depth of the target characteristic peak. Therefore, the peak shape characteristics and the peak distribution concentration characteristics of the target depth of the target characteristic peak are integrated to obtain the characteristic value of the target depth of the target characteristic peak. In an exemplary embodiment, a specific quantification method of the characteristic value is given as follows:
[0075]
[0076] Among them, S i,c A represents the characteristic value of the i-th characteristic peak at the c-th depth; i,c represents the peak area of the i-th characteristic peak at the c-th depth; W i,c represents the peak width of the i-th characteristic peak at the c-th depth; represents the peak shape characteristics of the i-th characteristic peak at the c-th depth; N represents the number of other signal points of the i-th characteristic peak at the c-th depth except the signal point corresponding to the peak; Y i,c,j represents the Raman intensity of the jth signal point of the i-th characteristic peak at the c-th depth; X i,c,j represents the Raman frequency shift of the jth signal point of the i-th characteristic peak at the c-th depth (the j-th signal point does not include the signal point corresponding to the peak); X i,c represents the Raman frequency shift corresponding to the peak of the i-th characteristic peak at the c-th depth; It represents the peak shape characteristics of the i-th characteristic peak at the c-th depth. In essence, it is the ratio of the Raman intensity of different signal points of the i-th characteristic peak at the c-th depth to the Raman frequency shift distance between the position corresponding to the peak. The closer to the peak position and the higher the Raman intensity, the more concentrated the distribution of the i-th characteristic peak at the c-th depth.
[0077] The larger the characteristic value of the target characteristic peak at the target depth, the more obvious the characteristics of the target characteristic peak at the target depth, and as the depth continues to increase, the characteristic value of the target characteristic peak will continue to decrease, thereby obtaining the characteristic value of each characteristic peak at each depth.
[0078] Step S2-2: Obtain the characteristic variation of each depth of each characteristic peak based on the difference in characteristic values of two adjacent depths of each characteristic peak.
[0079] Fruit and vegetable matrix components vary at different depths, potentially interfering with the Raman signal through absorption, scattering, or fluorescence, thereby affecting the intensity of the fungicide's characteristic peak. By collecting Raman spectra at different depths in fruits and vegetables and comparing the intensity changes of the fungicide's characteristic peak at each depth, we can determine whether the signal is affected by matrix interference. This allows us to determine the extent to which the fungicide's characteristic peak is affected by matrix components, analyze the impact of matrix interference, and improve the accuracy and reliability of fungicide residue determination.
[0080] Because the composition and structure of the fruit and vegetable matrix vary at different depths, the signal intensity of the fungicide's Raman signature peak may vary due to light scattering, absorption, or matrix interference when penetrating different depths. By comparing the changes in the characteristic values of the fungicide's signature peak at two adjacent depths, we can determine the characteristic variation of the peak at those two adjacent depths, and then determine the characteristic variation of the peak at each depth. This characteristic variation at each depth can be used to assess whether the fruit and vegetable matrix is interfering.
[0081] In an exemplary embodiment, a specific calculation method of the characteristic change is given as follows:
[0082]
[0083] Among them, B i,c It represents the characteristic change of the i-th characteristic peak at the c-th depth, which is essentially the characteristic change rate; S i,c+1 It represents the characteristic value of the i-th characteristic peak at the c+1-th depth (that is, the next depth after the c-th depth).
[0084] By the above method, the characteristic change rate of each depth of the i-th characteristic peak is obtained. It should be understood that, by the above method, the characteristic change rate of the i-th characteristic peak at the last depth is no longer obtained.
[0085] The layer-by-layer differences in the internal structure and composition of fruits and vegetables can lead to inconsistent effects of Raman excitation light on the fungicide's characteristic peaks at different depths. By analyzing the characteristic change rate of the fungicide's characteristic peak at each depth and integrating its continuous changes in depth, we can identify the changing trend of the characteristic peak with depth. Therefore, based on the characteristic changes in the Raman spectral data at each depth for each characteristic peak, we can determine the changing trend of the Raman signal intensity for each characteristic peak.
[0086] In an exemplary embodiment, Figure 5 As shown, a specific process for obtaining the Raman signal intensity variation trend is given below:
[0087] Step S2-3: Obtain the fluctuation degree of the characteristic change at each depth of each characteristic peak.
[0088] For the target characteristic peak, the degree of fluctuation of the characteristic change rate of each depth of the target characteristic peak is obtained. The degree of fluctuation represents the change of the characteristic change rate at each depth. The more consistent the change, that is, the more consistent the change trend, the higher the change trend of the Raman signal intensity of the target characteristic peak.
[0089] In an exemplary embodiment, a specific calculation method for the degree of fluctuation of the characteristic change rate of each depth of the target characteristic peak is given as follows:
[0090]
[0091] Among them, E i represents the degree of fluctuation of the characteristic change rate of each depth of the i-th characteristic peak; M represents the number of depths; B i Represents the average of the feature change rates of all depths of the i-th feature peak. The reason for taking M-1 data is that there is no feature change rate of the last layer, and the feature change rate of the last layer is not included in the calculation when calculating the average of the feature change rate.
[0092] Therefore, E i The calculation formula for is as follows: the absolute value of the difference between the characteristic change rate at each depth of the i-th characteristic peak and the average characteristic change rate is obtained, and then the average of the absolute values of the difference is calculated to obtain the degree of fluctuation of the characteristic change rate of the i-th characteristic peak. It should be understood that, as another embodiment, the variance of the characteristic change rate at each depth of the target characteristic peak can also be calculated as the degree of fluctuation.
[0093] Step S2-4: Obtain the Raman signal intensity variation trend of each characteristic peak according to the fluctuation degree.
[0094] Since the smaller the fluctuation of the characteristic change rate of each depth of the target characteristic peak, the higher the Raman signal intensity change trend of the target characteristic peak, the Raman signal intensity change trend is inversely proportional to the fluctuation degree. In an exemplary embodiment, the Raman signal intensity change trend is obtained by the following calculation method:
[0095] Q i =exp(-E i );
[0096] Among them, Q i represents the Raman signal intensity variation trend of the i-th characteristic peak, and exp is an exponential function with the natural constant e as the base. Through the above method, the Raman signal intensity variation trend of each characteristic peak is obtained.
[0097] Step S3: Obtain the degree of matrix interference influence on each depth of each characteristic peak based on the difference in trend change corresponding to all depths before and after a certain depth of each characteristic peak.
[0098] If the characteristic peak is unaffected by the fruit and vegetable matrix, its Raman signal intensity should gradually decrease with increasing depth, and the value of the Raman signal intensity change trend should be larger. Conversely, if the fungicide is affected by the matrix at a certain depth, the original Raman signal intensity change trend will be broken, and the value of the Raman signal intensity change trend will be smaller. Therefore, the Raman signal intensity change trend of the characteristic peak unaffected by the matrix can be compared to extract the trend change differences of the characteristic peak affected by the matrix.
[0099] Since the Raman signal intensity change trend is the change in the characteristic change rate reflected by the characteristic change rate of multiple consecutive depths, it is not a parameter that can be characterized by a single depth. Therefore, when obtaining any Raman signal intensity change trend, it is necessary to define two depths, one of which is the starting depth and the other is the cut-off depth. The Raman signal intensity change trend is the change in the characteristic change rate from the characteristic change rate of the starting depth to the characteristic change rate of the cut-off depth. Sort the depths from shallow to deep, namely: the first depth, the second depth, the third depth, and so on. For example: if the starting depth is the first depth and the cut-off depth is the fourth depth, then based on the characteristic change rates of the four depths from the first depth to the fourth depth, calculate the degree of fluctuation of the four characteristic change rates. In the calculation formula for the degree of fluctuation, B i,c The c in B are 1, 2, 3, and 4 respectively. i is the average value of the characteristic change rate of the four depths, thereby calculating the change trend of the Raman signal intensity for the four depths.
[0100] Therefore, the calculation method provided above can be used to obtain the Raman signal intensity change trend corresponding to any starting depth and cutoff depth of the target characteristic peak, and then obtain the difference between the Raman signal intensity change trend and the standard Raman signal intensity change trend corresponding to the starting depth and cutoff depth, specifically the absolute value of the difference, as the trend change difference of the target characteristic peak for the starting depth and cutoff depth, that is, the trend change difference represents the difference between the Raman signal intensity change trend of the characteristic peak and the standard Raman signal intensity change trend.
[0101] It should be understood that the standard Raman signal intensity variation trend is the Raman signal intensity variation trend of the characteristic peak of the fungicide that is not interfered with by the fruit and vegetable matrix. The standard Raman signal intensity variation trend is a known quantity obtained in advance. The standard Raman signal intensity variation trend of each characteristic peak of the corresponding fungicide that is not interfered with by the fruit and vegetable matrix can be obtained by experimental means in the laboratory. As a specific embodiment, Figure 6 As shown, the process of obtaining the standard Raman signal intensity variation trend can be:
[0102] Step S3-1: Acquire Raman spectrum data samples of multiple standard samples.
[0103] In the laboratory, multiple standard samples are obtained, wherein the standard samples are fruits and vegetables whose fruit and vegetable matrix does not interfere with the characteristic peaks of the fungicide, and the fruits and vegetables are the same as the above-mentioned measurement objects. For example, if the measurement object is citrus, the standard sample is also citrus. It should be understood that in the laboratory, in order to ensure that the fruit and vegetable matrix does not interfere with the characteristic peaks of the fungicide, the fungicide spraying method can be optimized to physically reduce matrix interference. For example, the fungicide atomizing nozzle can be replaced with an ultrasonic atomizing nozzle, and the droplet diameter is within the low interference range, such as 30-50μm, and the spray pressure is within the low interference range, such as 0.8MPa, etc. Furthermore, the fungicide can be sprayed within the optimal spraying time, such as spraying 7 days before picking, and immediately sent to the laboratory after spraying to reduce matrix interference. In addition, a film-forming agent can be added to form a 1-5μm isolation film on the surface of the fruit and vegetable without affecting the Raman spectral data to reduce matrix interference. Through the above-mentioned several improvement measures, the matrix interference is minimized to obtain multiple standard samples. The number of standard samples selected is set according to actual needs. In order to ensure data reliability, a larger number of standard samples is selected. The Raman spectrum of each standard sample is obtained and defined as a Raman spectrum data sample.
[0104] Step S3-2: Obtaining a Raman signal intensity variation trend of each characteristic peak of each standard sample according to characteristic variations of the Raman spectrum data samples at each depth of each characteristic peak of each standard sample.
[0105] The Raman signal intensity variation trend of each characteristic peak of each standard sample is obtained by adopting the method for obtaining the Raman signal intensity variation trend described above, based on the characteristic changes of the Raman spectrum data samples at each depth of each characteristic peak of each standard sample. That is, the Raman signal intensity variation trend of each characteristic peak of each standard sample for a certain starting depth and cutoff depth is obtained.
[0106] Step S3-3: Calculate the average of the Raman signal intensity variation trends of the same characteristic peak of all standard samples as the standard Raman signal intensity variation trend of the corresponding characteristic peak.
[0107] Since the Raman signal intensity variation trend corresponding to each standard sample is obtained for a certain starting depth and cutoff depth of the same characteristic peak, the average value of the Raman signal intensity variation trend corresponding to all standard samples for a certain starting depth and cutoff depth of the same characteristic peak is calculated. This average value is the standard Raman signal intensity variation trend corresponding to a certain starting depth and cutoff depth of a certain characteristic peak. In this way, the standard Raman signal intensity variation trend of each characteristic peak for each starting depth and cutoff depth is obtained. Each starting depth and cutoff depth represents all possible combinations of starting depths and cutoff depths. Obtaining the standard Raman signal intensity variation trend by calculating the average value can improve the reliability of the standard Raman signal intensity variation trend.
[0108] By adopting the above process, all possible combinations of starting depth and cutoff depth of each characteristic peak are traversed to obtain the standard Raman signal intensity variation trend of all possible combinations of starting depth and cutoff depth of each characteristic peak.
[0109] For a target characteristic peak, the greater the difference in the Raman signal intensity change trend of the target characteristic peak and the corresponding standard Raman signal intensity change trend, the more severe the interference of the target characteristic peak with the fruit and vegetable matrix. It should be understood that since the standard Raman signal intensity change trend corresponds to fruits and vegetables whose fruit and vegetable matrix does not interfere with the fungicide characteristic peak, if the difference in the Raman signal intensity change trend of a characteristic peak and the corresponding standard Raman signal intensity change trend is zero, it indicates that the fruit and vegetable matrix to be tested does not interfere with the fungicide characteristic peak. In this case, the subsequent data processing process provided in this embodiment is no longer performed, and the final fungicide residual amount is directly determined based on the collected fungicide characteristic peaks.
[0110] If the target characteristic peak of the fungicide is interfered with by the fruit and vegetable matrix at a certain depth, the more obvious the matrix components inside the fruit and vegetable, the deeper the depth of the fruit and vegetable, the stronger the degree of suppression of the characteristic performance of the interfered fungicide, that is, the smaller the trend change difference. Therefore, according to the difference in trend change differences at all depths before and after a certain depth of the target characteristic peak, that is, comparing the trend change differences before and after a certain depth, the degree of matrix interference influence on the target characteristic peak at that depth is obtained. In an exemplary embodiment, if Figure 7 As shown, a specific process for obtaining the degree of matrix interference is given as follows:
[0111] Step S3 - 4 : obtaining first trend change differences corresponding to all depths before the target depth of the target characteristic peak, and second trend change differences corresponding to all depths after the target depth of the target characteristic peak.
[0112] Obtain the trend change difference corresponding to all depths prior to the target depth of the target characteristic peak, which is defined as the first trend change difference. For example, if the target depth is the fifth depth, then obtain the Raman signal intensity change trend corresponding to the first to fourth depths of the target characteristic peak, as well as the standard Raman signal intensity change trend corresponding to the first to fourth depths of the target characteristic peak. Calculate the absolute value of the difference between the Raman signal intensity change trend and the standard Raman signal intensity change trend. The resulting trend change difference is the first trend change difference corresponding to all depths prior to the fifth depth of the target characteristic peak, which can also be expressed as the first trend change difference corresponding to the fifth depth of the target characteristic peak.
[0113] Similarly, the trend change difference corresponding to all depths after the target depth of the target characteristic peak is obtained, which is defined as the second trend change difference. For example, if the target depth is the fifth depth, then the Raman signal intensity change trend corresponding to the sixth depth to the last depth of the target characteristic peak is obtained (essentially, the Raman signal intensity change trend from the sixth depth to the second-to-last depth, the same applies below), as well as the standard Raman signal intensity change trend corresponding to the sixth depth to the last depth of the target characteristic peak. The absolute value of the difference between this Raman signal intensity change trend and the standard Raman signal intensity change trend is calculated. The resulting trend change difference is the second trend change difference corresponding to all depths after the fifth depth of the target characteristic peak, which can also be expressed as the second trend change difference corresponding to the fifth depth of the target characteristic peak.
[0114] Step S3-5: Obtain the relative deviation of the second trend change difference with respect to the first trend change difference to obtain the matrix interference influence degree of the target depth of the target characteristic peak.
[0115] The relative deviation of the second trend change difference relative to the first trend change difference indicates the proportion of the difference between the second trend change difference and the first trend change difference to the first trend change difference. The larger the relative deviation, the greater the difference between the second trend change difference and the first trend change difference, and the more serious the impact of the fungicide on the fruit and vegetable matrix interference at the target depth, that is, the higher the degree of matrix interference influence at the target depth. The calculation formula for the degree of matrix interference influence is as follows:
[0116]
[0117] Among them, G i,c represents the degree of matrix interference influence on the i-th characteristic peak at the c-th depth, that is, the degree of matrix interference influence on the i-th characteristic peak at the c-th depth; D′ i,c represents the first trend change difference of all depths before the cth depth of the i-th characteristic peak; D i,cIndicates the second trend change difference corresponding to all depths after the cth depth of the i-th characteristic peak.
[0118] It should be understood that the above calculation method cannot obtain the degree of matrix interference influence at the first depth and the degree of matrix interference influence at the last two depths. Accordingly, the initial fungicide residues at the first depth and the last two depths will not be corrected in the subsequent process.
[0119] Step S4: correcting the initial fungicide residue measured by the Raman spectrum data at each depth of each characteristic peak according to the degree of matrix interference and the characteristic changes at each depth of each characteristic peak.
[0120] Since the fruit and vegetable matrix has uneven interference effects on the Raman signal at different depths, the actual Raman intensity of the fungicide characteristic peak may deviate from its true concentration response. The specific impact of the matrix interference effect on the Raman signal intensity can be quantified by the degree of matrix interference effect on each characteristic peak at each depth, and the initial fungicide residue measured by the Raman spectral data can be corrected accordingly, thereby accurately reflecting the actual residual level of the fungicide at each depth. Therefore, according to the degree of matrix interference effect and the characteristic changes at each depth of the target characteristic peak, the initial fungicide residue measured by the Raman spectral data at each depth of the target characteristic peak is corrected. Figure 8 As shown, a specific correction process is given:
[0121] Step S4-1: The matrix interference effect degree and characteristic variation difference of the target depth of the target characteristic peak are integrated to obtain a correction coefficient of the target depth of the target characteristic peak.
[0122] Obtain the difference between the characteristic change rate of the target depth of the target characteristic peak and the standard characteristic change rate, specifically the absolute value of the difference between the characteristic change rate and the standard characteristic change rate, as the characteristic change difference of the target depth of the target characteristic peak. Among them, the standard characteristic change rate indicates that the detection object is fruits and vegetables whose fruit and vegetable matrix does not interfere with the fungicide characteristic peak. When performing Raman spectroscopy on such fruits and vegetables, the characteristic change rate of the target depth of the target characteristic peak obtained is used as the standard characteristic change rate of the target depth of the standard characteristic peak. In an exemplary embodiment, while obtaining the standard Raman signal intensity change trend of the target depth of the target characteristic peak, for the sake of correspondence, the target characteristic peak can be defined as the standard characteristic peak. Then, the standard characteristic change rate is the characteristic change rate of the target depth of the standard characteristic peak, that is, the characteristic change rate of the target depth of the target characteristic peak. Since the process of obtaining the characteristic change rate has been described above, it will not be repeated here.
[0123] The degree of matrix interference influence of the target depth of the target characteristic peak and the characteristic change difference of the target depth of the target characteristic peak are integrated to obtain a correction coefficient for the target depth of the target characteristic peak. The correction coefficient is proportional to the degree of matrix interference influence and the characteristic change difference. In an exemplary embodiment, a specific method for obtaining the correction coefficient is given as follows:
[0124] X i,c =1+G i,c ×ΔB i,c ;
[0125] Among them, X i,c represents the correction coefficient of the cth depth of the ith characteristic peak; ΔB i,c The absolute value of the difference between the characteristic change rate of the cth depth of the i-th characteristic peak and the standard characteristic change rate of the cth depth of the i-th characteristic peak. i,c ×ΔB i,c It represents the difference between the characteristic change rate of the i-th characteristic peak at the c-th depth and the standard characteristic change rate, combined with the degree of matrix interference, to obtain the adjustment range of the fungicide residue.
[0126] Step S4-2: Correcting the initial fungicide residue corresponding to the target depth of the target characteristic peak according to the correction coefficient.
[0127] It should be understood that a standard curve is established by fitting the relationship between the characteristic values of each characteristic peak at each depth and the fungicide concentration measured under the same experimental conditions for a fungicide sample with known fungicide concentration (the fitting relationship is usually a linear fitting relationship), and the characteristic values of the fungicide at each depth are mapped to the standard curve to obtain the initial fungicide residue at each depth of each characteristic peak.
[0128] The correction coefficient of the target depth of the target characteristic peak is multiplied by the initial fungicide residue at the target depth of the target characteristic peak. The result is the corrected fungicide residue at the target depth of the target characteristic peak. The calculation formula is as follows:
[0129] F i,c =L i,c ×X i,c ;
[0130] Among them, F i,c represents the residual amount of fungicide after the cth depth correction of the ith characteristic peak, L i,c represents the initial fungicide residue at the cth depth of the ith characteristic peak.
[0131] Step S5: The final fungicide residue of each characteristic peak is obtained based on the corrected fungicide residue at each depth of each characteristic peak and the degree of matrix interference.
[0132] Fusion of Raman signals at different depths helps overcome biases that can arise from single-layer measurements. While surface-level signals are strong, they are susceptible to the influence of concentrated surface residues. Deep-level signals, while closer to the overall distribution, are susceptible to interference from the fruit and vegetable matrix. By integrating multi-layer Raman information to determine the degree to which the Raman signals of fungicides at each depth are affected by interference from the fruit and vegetable matrix, the final fungicide residue is derived, enabling a more comprehensive assessment and more accurately reflecting the fungicide residue in the entire fruit and vegetable.
[0133] In an exemplary embodiment, a matrix interference influence degree sequence is obtained. The matrix interference influence degree sequence is composed of the matrix interference influence degrees at each depth of the target characteristic peak, and the matrix interference influence degrees are sorted from shallow to deep. A fungicide residue sequence is obtained. The fungicide residue sequence is composed of the corrected fungicide residues at each depth of the target characteristic peak, and the corrected fungicide residues are also sorted from shallow to deep.
[0134] The matrix interference influence degree sequence of the target characteristic peak is used as the weight to obtain the weighted average of the fungicide residue sequence of the target characteristic peak, which is used as the final fungicide residue of the target characteristic peak. The calculation formula is as follows:
[0135]
[0136] Among them, Z i represents the final fungicide residue of the i-th characteristic peak. Combined with the corrected fungicide residues at each depth, the distribution characteristics of fungicides at different depths of fruits and vegetables can be accurately identified, further improving the accuracy of fungicide residue determination.
[0137] It should be understood that since the fungicide residues at the first depth and the last two depths are not included in the correction, the final fungicide residues calculated above are only for the final fungicide residues from the second depth to the third to the last depth. Then, they are fused with the initial fungicide residues at the first depth and the last two depths to obtain the final fungicide residues for all depths of the characteristic peak. The fusion method is: for the i-th characteristic peak, calculate Z i The final fungicide residue of the i-th characteristic peak is obtained by calculating the average of the initial fungicide residue at the first depth of the i-th characteristic peak, the initial fungicide residue at the second to last depth of the i-th characteristic peak, and the initial fungicide residue at the penultimate depth of the i-th characteristic peak.
[0138] Using this method, the final fungicide residue for each characteristic peak of the fungicide is obtained. In subsequent applications, the obtained final fungicide residue for each characteristic peak can be used to generate a fungicide residue report. In addition to the final fungicide residue for each characteristic peak, it can also include Raman spectral data at various depths of each characteristic peak, serving as a technical basis for food safety traceability.
[0139] This embodiment further provides a system for accurately measuring the residual amount of a fruit and vegetable fungicide, comprising: a memory and a processor; the memory is connected to the processor, the memory being used to store program instructions; and the processor being used to implement the steps of the above-mentioned embodiment of the method for accurately measuring the residual amount of a fruit and vegetable fungicide when the program instructions are executed.
[0140] In an exemplary embodiment, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for accurately determining the residual amount of a fungicide for fruits and vegetables.
[0141] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0142] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.
Claims
1. A method for accurately determining the residual amount of fungicide in fruits and vegetables, characterized by: include: Obtain Raman spectral data of fruits and vegetables at different depths; According to the characteristic changes of the Raman spectrum data of each characteristic peak at each depth, the Raman signal intensity change trend of each characteristic peak is obtained; The degree of influence of matrix interference on each depth of each characteristic peak is obtained based on the difference in trend change differences corresponding to all depths before and after a certain depth of each characteristic peak; the trend change difference represents the difference between the Raman signal intensity change trend of the characteristic peak and the standard Raman signal intensity change trend; Correcting the initial fungicide residue measured by the Raman spectrum data at each depth of each characteristic peak according to the degree of matrix interference and the characteristic changes at each depth of each characteristic peak; According to the corrected fungicide residue at each depth of each characteristic peak and the degree of matrix interference, the final fungicide residue of each characteristic peak is obtained.
2. The method for accurately determining the residual amount of a fruit and vegetable fungicide according to claim 1, wherein: The process of acquiring the feature change includes: Determining the characteristic value of each characteristic peak at each depth from the Raman spectrum data of each characteristic peak at each depth; The characteristic variation of each characteristic peak at each depth is obtained from the difference in characteristic values of two adjacent depths of each characteristic peak.
3. The method for accurately determining the residual amount of a fruit and vegetable fungicide according to claim 2, wherein: The process of obtaining the characteristic value includes: Obtaining a peak shape characteristic according to the peak area and peak width of a target characteristic peak at a target depth; the target characteristic peak is any characteristic peak, and the target depth is any depth; the peak shape characteristic is proportional to the peak area and inversely proportional to the peak width; Obtaining a Raman frequency shift difference between each signal point and a signal point corresponding to the peak at a target depth of the target characteristic peak, and combining the Raman intensity of each signal point at the target depth of the target characteristic peak to obtain a peak distribution concentration feature at the target depth of the target characteristic peak; the peak distribution concentration feature is inversely proportional to the Raman frequency shift difference and directly proportional to the Raman intensity; The peak shape feature and the peak distribution concentration feature are integrated to obtain a characteristic value of a target depth of a target characteristic peak.
4. The method for accurately determining the residual amount of a fruit and vegetable fungicide according to claim 1, wherein: The process of obtaining the Raman signal intensity variation trend includes: Obtaining the degree of fluctuation of the characteristic change at each depth of each characteristic peak; According to the fluctuation degree, a Raman signal intensity variation trend of each characteristic peak is obtained, and the Raman signal intensity variation trend is inversely proportional to the fluctuation degree.
5. The method for accurately determining the residual amount of a fruit and vegetable fungicide according to claim 1, wherein: The process of obtaining the standard Raman signal intensity variation trend includes: Obtaining Raman spectral data samples of a plurality of standard samples; the standard samples are fruits and vegetables whose matrices do not interfere with the characteristic peaks of the fungicide; According to the characteristic changes of the Raman spectrum data samples at each depth of each characteristic peak of each standard sample, the Raman signal intensity change trend of each characteristic peak of each standard sample is obtained; The average of the Raman signal intensity variation trends of the same characteristic peak of all standard samples is calculated as the standard Raman signal intensity variation trend of the corresponding characteristic peak.
6. The method for accurately determining the residual amount of a fruit and vegetable fungicide according to claim 1, wherein: The process of obtaining the degree of matrix interference influence includes: Obtaining a first trend change difference corresponding to all depths before a target depth of a target characteristic peak, and a second trend change difference corresponding to all depths after the target depth of the target characteristic peak; the target characteristic peak is any characteristic peak, and the target depth is any depth; The relative deviation of the second trend change difference with respect to the first trend change difference is obtained to obtain the degree of matrix interference influence on the target depth of the target characteristic peak.
7. The method for accurately determining the residual amount of a fungicide in fruits and vegetables according to claim 1, wherein: The process of correcting the initial fungicide residue measured by the Raman spectrum data at each depth of each characteristic peak includes: The target depth correction coefficient of the target characteristic peak is obtained by fusing the matrix interference influence degree and the characteristic change difference of the target characteristic peak; the correction coefficient is proportional to the matrix interference influence degree and the characteristic change difference; the target characteristic peak is any characteristic peak, and the target depth is any depth; the characteristic change difference is the difference between the characteristic change of the target depth of the target characteristic peak and the standard characteristic change; The initial fungicide residue corresponding to the target depth of the target characteristic peak is corrected according to the correction coefficient.
8. The method for accurately determining the residual amount of a fruit and vegetable fungicide according to claim 7, wherein: The process of obtaining the standard feature change includes: A standard characteristic peak is obtained, where the standard characteristic peak is a characteristic peak corresponding to the standard Raman signal intensity change trend; and the standard characteristic change is a characteristic change of a target depth of the standard characteristic peak.
9. The method for accurately determining the residual amount of a fruit and vegetable fungicide according to claim 1, wherein: The process of obtaining the terminal fungicide residue includes: Taking the matrix interference influence degree sequence of the target characteristic peak as the weight, the weighted average value of the fungicide residue sequence of the target characteristic peak is obtained as the final fungicide residue of the target characteristic peak; the target characteristic peak is any characteristic peak, the matrix interference influence degree sequence is composed of the matrix interference influence degrees of each depth of the target characteristic peak, and the fungicide residue sequence is composed of the corrected fungicide residues at each depth of the target characteristic peak.
10. A system for accurately measuring the residual amount of fungicide in fruits and vegetables, comprising: Memory and processor; The memory is connected to the processor; The memory is used to store program instructions; The processor is configured to implement the method for accurately determining the residual amount of a fruit and vegetable fungicide according to any one of claims 1 to 9 when the program instructions are executed.