A quality detection method for rice processing

Abstract

The invention discloses a quality detection method for rice processing, and relates to the technical field of quality detection of rice processing. The method is implemented by using a spectral data acquisition module, a data processing module, a comprehensive comparison difference evaluation module, a feedback adjustment module and a quality control decision module. The spectral data acquisition module is used to collect spectral data of rice samples of the same batch. The data processing module is used to pre-process the collected spectral data. The comprehensive comparison difference evaluation module is used to output a multi-wavelength spectral difference index and a comprehensive quality evaluation index of the rice. The multi-wavelength spectral difference index and the comprehensive quality evaluation index are input into a feedback adjustment module, and a reflectivity after adjustment of a third wavelength s is output. Based on the influence of the reflectivity after adjustment of the third wavelength s on the spectral characteristics of the rice surface, subsequent processing of the rice is optimized or quality controlled.

Description

Technical Field

[0001] The invention relates to the technical field of quality detection of rice processing, in particular to a quality detection method for rice processing. Background Art

[0002] Rice, also known as rice, is a food made from rice seeds after processes such as cleaning, husking, milling, and finished product arrangement. Rice contains nearly 64% of the nutrients in rice and more than 90% of the nutrients required by the human body. It can be divided into three types: indica rice, japonica rice, and glutinous rice. In the rice processing process, quality inspection is a key link to ensure product quality and consumer satisfaction.

[0003] At present, the detection of rice processing precision at home and abroad generally adopts the method of manual judgment. This method has the shortcomings of large subjective deviation and difficulty in real-time online detection. With the development of spectral analysis, image processing and other technologies, more and more researchers have begun to explore the application of these technologies in the quality detection of rice processing. Specifically, the method of manual judgment is very subjective, and the judgment results of different people may be inconsistent, and the subjective deviation is large; the traditional method usually requires sampling and laboratory analysis, which cannot reflect the quality changes in the production process in real time; the spectral data volume is large and the information is rich, but there is also noise and redundant information, which requires effective data processing methods. Summary of the invention

[0004] The object of the present invention is to provide a quality detection method for rice processing, which solves the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solution, using:

[0006] It is realized by using spectral data acquisition module, data processing module, comprehensive comparison difference evaluation module, feedback adjustment module and quality control decision module;

[0007] The specific implementation steps are as follows:

[0008] Spectral data acquisition: spectral data of the same batch of rice samples are collected through the spectral data acquisition module;

[0009] Data processing: pre-process the collected spectral data through the data processing module;

[0010] Rice processing optimization and feedback adjustment: output the multi-wavelength spectrum difference index BC and the comprehensive quality evaluation index ZP of rice through the comprehensive comparison difference evaluation module, input the multi-wavelength spectrum difference index BC and the comprehensive quality evaluation index ZP into the feedback adjustment module, output the reflectivity TF after the third wavelength s is adjusted, and optimize or control the quality of the subsequent processing of rice based on the influence of the reflectivity TF after the third wavelength s is adjusted on the spectral characteristics of the rice surface;

[0011] The comprehensive comparison difference assessment module includes a rice surface spectrum difference unit under multiple wavelengths and a rice comprehensive quality assessment unit;

[0012] The feedback adjustment module includes a rice surface spectrum analysis unit after feedback adjustment.

[0013] Optionally, the difference manifestation process of the rice surface spectrum difference unit under the multi-wavelength is as follows:

[0014]

[0015] in:

[0016] BC is the multi-wavelength spectral difference index;

[0017] m is the total number of wavelengths;

[0018] y is the first wavelength y, and y represents the multi-wavelength consideration interval from 1 to m;

[0019] F y is the reflectivity at the first wavelength y;

[0020] BP is the average index of multi-wavelength reflectivity;

[0021] FB is the multi-wavelength reflectivity standard deviation index;

[0022] The reflectivity F at the first wavelength y processed by the data processing module y , the multi-wavelength reflectivity average index BP and the multi-wavelength reflectivity standard deviation index FB are input into the rice surface spectrum difference unit under multi-wavelength to obtain the multi-wavelength spectrum difference index BC for evaluating the complexity and uniformity of rice surface characteristics.

[0023] Optionally, the evaluation process of the rice comprehensive quality evaluation unit is as follows:

[0024] ZP=BC*(α*log10(F x +ε)+τ);

[0025] in:

[0026] ZP is the comprehensive quality assessment index;

[0027] BC is the multi-wavelength spectral difference index;

[0028] F x is the reflectivity at the second wavelength x;

[0029] α is the logarithmic transformation adjustment coefficient, which is used to adjust F x slope, translation, and overall scaling of logarithmic transformations;

[0030] ε is the avoidance coefficient, which is used to avoid the negative infinity or zero value problem in the logarithmic transformation;

[0031] τ is the overall level adjustment coefficient, and τ acts on the degree of adjustment of the overall level of ZP;

[0032] The multi-wavelength spectral difference index BC obtained by the rice surface spectral difference unit under multi-wavelength is input into the rice comprehensive quality assessment unit to obtain a comprehensive quality assessment index ZP that can identify the differences between different batches or varieties and conduct a comprehensive assessment of rice quality.

[0033] Optionally, the analysis process of the rice surface spectrum analysis unit after the feedback adjustment is as follows:

[0034]

[0035] in:

[0036] TF is the reflectivity after adjustment at the third wavelength s;

[0037] F s is the spectral characteristic index at the third wavelength s;

[0038] δ is the adjustment factor, which reflects the influence of ZP on the reflectivity adjustment control;

[0039] ZP is the comprehensive quality assessment index;

[0040] MP is the comprehensive quality assessment target index;

[0041] DB is the maximum index of comprehensive quality assessment, XB is the minimum index of comprehensive quality assessment, DB and XB are the maximum and minimum values ​​of the same batch or the same ZP standard;

[0042] The comprehensive quality evaluation index ZP obtained from the rice comprehensive quality evaluation unit and the reflectance F after adjustment at the third wavelength s are used. s , comprehensive quality assessment target index MP, comprehensive quality assessment maximum index DB and comprehensive quality assessment minimum index XB are input into the rice surface spectral analysis unit after feedback adjustment to obtain the adjusted reflectance TF of the third wavelength s reflecting the influence of the comprehensive quality on the spectral characteristics.

[0043] Optionally, based on the adjusted reflectivity TF of the third wavelength s, the feedback adjustment analysis of the quality control decision module is as follows:

[0044] S1, the reflectivity TF of the third wavelength s adjusted by the rice surface spectrum analysis unit after feedback adjustment is used as the reflectivity F of the first wavelength y in the rice surface spectrum difference unit under the next round of multi-wavelengths y Replace the data, calculate, and then output the new multi-wavelength spectral difference index BC new ;

[0045] S2, if the new multi-wavelength spectral difference index BC new , compared with the value of the multi-wavelength spectral difference index BC, there is a more improved optimization trend, then it is considered that the control optimization measures for rice processing quality are effective, and in the subsequent processing process, the same accuracy of processing optimization or quality control should be maintained;

[0046] S3, if the new multi-wavelength spectral difference index BC new , compared with the value of the multi-wavelength spectral difference index BC, there is a lower improvement trend, which means that the control and optimization measures for rice processing quality are effective, but not strong enough. In the subsequent processing, processing optimization or quality control should be strengthened;

[0047] S4, if the new multi-wavelength spectral difference index BC new , compared with the value of the multi-wavelength spectral difference index BC, no improvement trend is produced, then the control optimization measures for rice processing quality are considered ineffective. In the subsequent processing process, when the processing optimization strategy or quality control method is changed, it is repeated several times until an effective improvement trend occurs.

[0048] Optionally, the reflectivity F at the first wavelength y is y The complexity and uniformity of the rice surface characteristics are reflected by calculating the difference in spectral reflectance at different wavelengths, which is used as the basic data of the rice surface spectral difference unit at multiple wavelengths; the reflectance F at the second wavelength x is x As an input variable in the rice comprehensive quality assessment unit; the reflectivity F after the third wavelength s is adjusted s is the spectral characteristic value after some preprocessing or feature extraction algorithm, wherein the reflectivity F at the first wavelength y y and the reflectivity F at the second wavelength x x Directly represents the original spectral reflectance, the third wavelength s adjusted reflectance F s After preprocessing or feature extraction algorithm.

[0049] Optionally, the data preprocessing of the data processing module includes preprocessing operations of cleaning, denoising and normalizing the spectral data.

[0050] Optionally, the equipment used in the spectral data acquisition module includes a spectrometer, and the equipment used in the data processing module includes a computer, data processing software and quality control equipment.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] 1. The present invention can more accurately reflect the quality of rice by quantifying spectral differences and comprehensively evaluating quality, thereby reducing the subjective bias of manual judgment and improving the accuracy of reflecting the quality of rice.

[0053] 2. The spectral analysis technology of the present invention has the advantages of rapid, non-contact and non-destructive detection, and can realize real-time online quality detection during rice processing, and utilizes the rapidity and non-contact characteristics of spectral analysis technology to realize real-time online quality detection.

[0054] 3. The present invention utilizes the application of the rice surface spectral analysis unit after feedback adjustment to optimize the preprocessing and feature extraction process of spectral data, remove noise and redundant information, and improve the efficiency and accuracy of data processing.

[0055] 4. The present invention utilizes an automated quality inspection system, which can reduce manual intervention, improve production efficiency and product quality stability, and thus promote automated production. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 The method flow chart of the quality detection method of rice processing is as follows;

[0057] Figure 2 The overall structure diagram of the quality detection method for rice processing is shown in FIG.

[0058] Figure 3 Schematic diagram of the structure of the comprehensive comparison difference evaluation module of the present invention;

[0059] Figure 4 It is a structural diagram of the feedback adjustment module of the present invention. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] The quality detection method for rice processing is different from the existing quality detection method for rice processing. The existing quality detection method for rice processing has the disadvantages of large subjective deviation and difficulty in real-time online detection. Specifically, the method of manual judgment is very subjective, and the judgment results of different people may be inconsistent, and the subjective deviation is large. The traditional method usually requires sampling and laboratory analysis, which cannot reflect the quality changes in the production process in real time. The spectral data volume is large and the information is rich, but there is also noise and redundant information, which loses the effectiveness of the data processing method. The algorithm unit can more accurately reflect the quality status of rice by quantifying spectral differences and comprehensively evaluating the quality, and realizes real-time online quality detection by utilizing the rapidity and non-contact characteristics of spectral analysis technology. The application of the rice surface spectral analysis unit after feedback adjustment can optimize the preprocessing and feature extraction process of spectral data, remove noise and redundant information, and improve the efficiency and accuracy of data processing, thereby reducing manual intervention, improving production efficiency and product quality stability, and promoting automated production.

[0062] For example, see Figures 1 to 4 ,This implementation provides a quality detection method for rice processing, which is implemented by a spectral data acquisition module, a data processing module, a comprehensive comparison and difference evaluation module, a feedback adjustment module, and a quality control decision module;

[0063] The specific implementation steps are as follows:

[0064] Spectral data acquisition: spectral data of the same batch of rice samples are collected through the spectral data acquisition module;

[0065] Data processing: pre-process the collected spectral data through the data processing module;

[0066] Rice processing optimization and feedback adjustment: output the multi-wavelength spectrum difference index BC and the comprehensive quality evaluation index ZP of rice through the comprehensive comparison difference evaluation module, input the multi-wavelength spectrum difference index BC and the comprehensive quality evaluation index ZP into the feedback adjustment module, output the reflectivity TF after the third wavelength s is adjusted, and optimize or control the quality of the subsequent processing of rice based on the influence of the reflectivity TF after the third wavelength s is adjusted on the spectral characteristics of the rice surface;

[0067] The comprehensive comparison difference evaluation module includes a rice surface spectrum difference unit under multiple wavelengths and a rice comprehensive quality evaluation unit;

[0068] The feedback adjustment module includes a rice surface spectrum analysis unit after feedback adjustment.

[0069] In this embodiment, the system has significant beneficial effects in improving detection accuracy, realizing real-time online detection, optimizing data processing and promoting automated production through the mutual cooperation of three algorithm units, and can effectively solve the shortcomings of the prior art. However, it should be noted that the rice surface spectral analysis unit after feedback adjustment needs to be combined with specific technical implementation and algorithm optimization to play its role in practical applications. Combining the three operation results of BC, ZP and TF, a high-precision quality detection method for rice processing that promotes automated production can be formed. This method can reduce manual intervention, improve production efficiency and product quality stability. BC is a multi-wavelength spectral difference index. A high multi-wavelength spectral difference index BC usually indicates that there is a large spectral difference on the rice surface. The reason is caused by factors such as impurities, uneven moisture or variety differences. The multi-wavelength spectral difference index BC is used to quantify the difference in spectral reflectance of rice at different wavelengths. This indicator can reflect the surface characteristics of rice. The complexity and uniformity of the rice are taken into consideration. ZP is a comprehensive quality evaluation index, which combines the multi-wavelength spectral difference index BC and the reflectance at a specific wavelength. The evaluation is carried out by introducing the logarithmic transformation adjustment coefficient α. The overall level adjustment coefficient τ is weighted and transformed to provide a more comprehensive rice quality. This model can comprehensively consider spectral differences and key quality indicators, such as moisture content and protein content. TF is the reflectance after adjustment of the third wavelength s. It feeds back the result value of the comprehensive quality evaluation index ZP to the original spectral characteristics, and corrects the spectral reflectance by adjusting the factor δ. This process aims to make the spectral characteristics closer to the characteristic performance under the target quality level, thereby improving the quality consistency and stability of rice. The calculation results of TF can also affect the feedback to the next round of BC and ZP calculations, so that the three algorithms of this system have a high correlation and entanglement, so that the overall algorithm system can be closer to reality by automatic feedback and optimization according to actual conditions.

[0070] See also Figures 1 to 4 , the difference of the spectrum difference unit of rice surface under multiple wavelengths is as follows:

[0071]

[0072]

[0073] in:

[0074] BC is the multi-wavelength spectral difference index;

[0075] m is the total number of wavelengths;

[0076] y is the first wavelength y, and y represents the multi-wavelength consideration interval from 1 to m;

[0077] F y is the reflectivity at the first wavelength y;

[0078] BP is the average index of multi-wavelength reflectivity;

[0079] FB is the multi-wavelength reflectivity standard deviation index;

[0080] The reflectivity F at the first wavelength y processed by the data processing module y , the multi-wavelength reflectivity average index BP and the multi-wavelength reflectivity standard deviation index FB are input into the rice surface spectrum difference unit under multi-wavelength to obtain the multi-wavelength spectrum difference index BC for evaluating the complexity and uniformity of rice surface characteristics.

[0081] In this embodiment:

[0082] First, the multi-wavelength reflectivity average index BP and the multi-wavelength reflectivity standard deviation index FB in this algorithm unit represent the average value and standard deviation of reflectivity at all wavelengths, respectively, which are used to normalize the reflectivity difference at different wavelengths, so that the multi-wavelength spectral difference index BC has better comparability, and the two are adopted and The calculation formula can calculate a more accurate average value and standard deviation, and the reflectivity F at the first wavelength y y , is the spectral reflectance of rice measured at a specific wavelength y, which is the result of direct experimental measurement. It is input into the rice surface spectral difference unit at multiple wavelengths to obtain the multi-wavelength spectral difference index BC used to evaluate the complexity and uniformity of rice surface characteristics;

[0083] This algorithm unit is to quantify the difference in the spectral reflectance of rice at different wavelengths, that is, the result value of the rice surface spectral difference unit at multiple wavelengths, the multi-wavelength spectral difference index BC. This difference is caused by many factors, including impurities on the rice surface, uneven moisture distribution, variety differences or changes in the processing process. The larger the value of the multi-wavelength spectral difference index BC, the more significant the difference in the spectral characteristics of rice at different wavelengths, which in turn implies that there are problems with the quality or consistency of the rice. The multi-wavelength spectral difference index BC can be used as a preliminary screening tool to quickly identify rice samples with abnormal spectral characteristics. During the processing, the multi-wavelength spectral difference index BC can help monitor the quality stability of rice, timely discover and adjust process parameters that may cause spectral differences, and achieve the purpose of quality control. For different varieties of rice, the multi-wavelength spectral difference index BC can show different characteristic values, which is helpful for variety differentiation and authentication, and provides characteristic spectral difference indicators for different varieties of rice.

[0084] See also Figures 1 to 4 , the evaluation process of the rice comprehensive quality evaluation unit is as follows:

[0085] ZP=BC*(α*log10(F x +ε)+τ);

[0086] in:

[0087] ZP is the comprehensive quality assessment index;

[0088] BC is the multi-wavelength spectral difference index;

[0089] F x is the reflectivity at the second wavelength x;

[0090] α is the logarithmic transformation adjustment coefficient, which is used to adjust F x slope, translation, and overall scaling of logarithmic transformations;

[0091] ε is the avoidance coefficient, which is used to avoid the negative infinity or zero value problem in the logarithmic transformation;

[0092] τ is the overall level adjustment coefficient, and τ acts on the degree of adjustment of the overall level of ZP;

[0093] The multi-wavelength spectral difference index BC obtained by the rice surface spectral difference unit under multi-wavelength is input into the rice comprehensive quality assessment unit to obtain a comprehensive quality assessment index ZP that can identify the differences between different batches or varieties and conduct a comprehensive assessment of rice quality.

[0094] In this embodiment, the logarithmic transformation adjustment coefficient α, the avoidance coefficient ε and the overall level adjustment coefficient τ are all obtained by fitting the experimental data and are used to adjust the reflectivity F at the second wavelength x. x The slope, translation, and overall scaling of the logarithmic transformation. Specifically, the logarithmic transformation adjustment factor α controls the reflectance F at the second wavelength x. x The avoidance coefficient ε is used to avoid the problem of negative infinity or zero value in logarithmic transformation, and the overall level adjustment coefficient τ is used to adjust the overall level of the comprehensive quality evaluation index ZP. The specific values ​​of these coefficients depend on the quality of experimental data and the choice of fitting method. By introducing the coefficient logarithmic transformation adjustment coefficient α, avoidance coefficient ε and overall level adjustment coefficient τ, as well as the reflectivity F at the second wavelength x x , reflectivity F at the second wavelength x x Represents the sensitive wavelengths of key quality indicators, including moisture or protein content, providing a more comprehensive quality assessment by combining spectral differences with key quality indicators, helping to identify differences between different batches or varieties;

[0095] The calculation of the comprehensive quality assessment index ZP combines the multi-wavelength spectral difference index BC and the reflectivity at a specific wavelength and the reflectivity F at a second wavelength x. x, aims to provide a more comprehensive rice quality assessment. By introducing the coefficient logarithmic transformation adjustment coefficient α, the avoidance coefficient ε and the overall level adjustment coefficient τ, the comprehensive quality assessment index ZP can adjust the contribution of different factors to the comprehensive quality assessment according to the experimental data, making the assessment results more accurate and reliable.

[0096] The comprehensive quality assessment index ZP takes into account spectral differences and key quality indicators, and can reflect the quality status of rice more comprehensively. It provides a scientific basis for quality control and decision-making in the processing process, including adjusting processing parameters and screening unqualified products. Finally, by setting unified evaluation standards and target values, it helps to achieve standardization and traceability of rice quality.

[0097] See also Figures 1 to 4 , the analysis process of the rice surface spectrum analysis unit after feedback adjustment is as follows:

[0098]

[0099] in:

[0100] TF is the reflectivity after adjustment at the third wavelength s;

[0101] F s is the spectral characteristic index at the third wavelength s;

[0102] δ is the adjustment factor, which reflects the influence of ZP on the reflectivity adjustment control;

[0103] ZP is the comprehensive quality assessment index;

[0104] MP is the comprehensive quality assessment target index;

[0105] DB is the maximum index of comprehensive quality assessment, XB is the minimum index of comprehensive quality assessment, DB and XB are the maximum and minimum values ​​of the same batch or the same ZP standard;

[0106] The comprehensive quality evaluation index ZP obtained from the rice comprehensive quality evaluation unit and the reflectance F after adjustment at the third wavelength s are used. s , comprehensive quality assessment target index MP, comprehensive quality assessment maximum index DB and comprehensive quality assessment minimum index XB are input into the rice surface spectral analysis unit after feedback adjustment to obtain the adjusted reflectance TF of the third wavelength s reflecting the influence of the comprehensive quality on the spectral characteristics.

[0107] In this embodiment, the algorithm unit is first based on the comprehensive quality assessment index ZP obtained in the rice comprehensive quality assessment unit, as well as the comprehensive quality assessment target index MP, the comprehensive quality assessment maximum index DB and the comprehensive quality assessment minimum index XB, wherein the comprehensive quality assessment target index MP is used to compare with the current comprehensive quality assessment index ZP, the comprehensive quality assessment maximum index DB and the comprehensive quality assessment minimum index XB belong to the maximum value and the minimum value of the comprehensive quality assessment index ZP in the same batch or standard, and are used for normalization adjustment, and the reflectance F after the third wavelength s is adjusted s , at wavelength F s The adjusted reflectivity is obtained by: wherein s can be any wavelength considered, and then the output third wavelength s adjusted reflectivity TF is obtained.

[0108] In this algorithm unit, the result of the comprehensive quality assessment index ZP is fed back to the original spectral characteristics through the calculation of the reflectance TF after adjustment of the third wavelength s. The spectral reflectance is corrected by adjusting the difference between the adjustment factor δ and the normalized comprehensive quality assessment index ZP. This correction aims to make the spectral characteristics closer to the characteristic performance under the target quality level.

[0109] The results of the third wavelength s-adjusted reflectance TF can be used to guide parameter adjustments during the processing process to optimize the final quality of rice and ensure optimized processing of rice; by continuously monitoring and adjusting the spectral characteristics, the stability and consistency of rice quality can be maintained and the quality of rice can be controlled; the third wavelength s-adjusted reflectance TF provides a new perspective and method for scientific research related to rice quality.

[0110] See also Figures 1 to 4 , based on the adjusted reflectivity TF of the third wavelength s, the feedback adjustment analysis of the quality control decision module is as follows:

[0111] S1, the reflectivity TF of the third wavelength s adjusted by the rice surface spectrum analysis unit after feedback adjustment is used as the reflectivity F of the first wavelength y in the rice surface spectrum difference unit under the next round of multi-wavelengths y Replace the data, calculate, and then output the new multi-wavelength spectral difference index BC new ;

[0112] S2, if the new multi-wavelength spectral difference index BC new , compared with the value of the multi-wavelength spectral difference index BC, there is a more improved optimization trend, then it is considered that the control optimization measures for rice processing quality are effective, and in the subsequent processing process, the same accuracy of processing optimization or quality control should be maintained;

[0113] S3, if the new multi-wavelength spectral difference index BC new, compared with the value of the multi-wavelength spectral difference index BC, there is a lower improvement trend, which means that the control and optimization measures for rice processing quality are effective, but not strong enough. In the subsequent processing, processing optimization or quality control should be strengthened;

[0114] S4, if the new multi-wavelength spectral difference index BC new , compared with the value of the multi-wavelength spectral difference index BC, no improvement trend is produced, then the control optimization measures for rice processing quality are considered invalid. In the subsequent processing, when the processing optimization strategy or quality control method is changed, it is repeated many times until an effective improvement trend occurs;

[0115] The reflectivity at the first wavelength y reflects the complexity and uniformity of the rice surface characteristics by calculating the difference in spectral reflectivity at different wavelengths, and serves as the basic data of the rice surface spectral difference unit at multiple wavelengths; the reflectivity at the second wavelength x serves as an input variable in the rice comprehensive quality evaluation unit; the adjusted reflectivity at the third wavelength s is a spectral characteristic value processed by preprocessing or feature extraction algorithm, wherein the reflectivity at the first wavelength y and the reflectivity at the second wavelength x directly represent the original spectral reflectivity, and the adjusted reflectivity at the third wavelength s has been processed by preprocessing or feature extraction algorithm.

[0116] In this embodiment, based on the result of the reflectance TF after adjustment of the third wavelength s, the algorithm unit can enable the processing enterprise to continuously optimize the production process and reduce the factors that cause abnormal spectral characteristics, thereby improving the overall quality of rice. This improvement may be reflected in the multi-wavelength spectral difference index BC value of subsequent batches of rice, which is manifested as a reduction in spectral differences and an improvement in consistency. For breeders, the result of the reflectance TF after adjustment of the third wavelength s can help identify rice varieties with better spectral characteristics, and then optimize the variety characteristics through genetic improvement. This optimization will also be reflected in the multi-wavelength spectral difference index BC value and other quality indicators related thereto. With the improvement of processing quality and the optimization of varieties, consumers will be able to purchase higher quality rice products. The positive changes in market feedback may further promote processing enterprises and breeders to continue to improve and optimize rice quality.

[0117] In summary, in the specific implementation process, the continuously optimized and adjusted reflectivity TF of the third wavelength s will bring different new multi-wavelength spectral difference indexes BC. new , and the multi-wavelength spectral difference index BC new By comparing with the previous multi-wavelength spectral difference index BC value, a more optimized rice quality control method will be obtained, which will help identify rice varieties with better spectral characteristics.

[0118] Reflectivity F at the first wavelength yy As the basic data for calculating the multi-wavelength spectral difference index BC, the reflectivity F at the first wavelength y y Directly affects the value of the multi-wavelength spectral difference index BC, which in turn reflects the quality characteristics of rice. As an input variable in the comprehensive quality evaluation index ZP model, the reflectance F at the second wavelength x x It is very important to evaluate the comprehensive quality of rice. The third wavelength s adjusted reflectance TF module uses the third wavelength s adjusted reflectance F s To optimize the preprocessing and feature extraction process of spectral data, remove noise and redundant information, and improve the efficiency and accuracy of data processing. Ultimately, these optimized eigenvalues ​​will be used for further quality assessment or model training.

[0119] For example 2, please refer to Figures 1 to 4 ,The data preprocessing of the data processing module includes the ,preprocessing operations of cleaning, denoising and normalization of the spectral ,data.

[0120] The equipment used in the spectrum data acquisition module includes a spectrometer, and the equipment used in the data processing module includes a computer, data processing software and quality control equipment.

[0121] In this embodiment, the spectrometer used in the spectral data acquisition module is used to collect spectral data of rice samples, the computer and data processing software are used for data processing, analysis and result display, and the quality control equipment can screen and remove impurities from the rice according to the quality evaluation results.

[0122] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A quality detection method for rice processing, characterized in that, The specific implementation steps are as follows: Spectral data collection: Spectral data of the same batch of rice samples are collected through the spectral data collection module; Data processing: pre-process the collected spectral data through the data processing module; Rice processing optimization and feedback adjustment: The multi-wavelength spectrum difference index BC and the comprehensive quality evaluation index ZP of rice are output through the comprehensive comparison difference evaluation module, the multi-wavelength spectrum difference index BC and the comprehensive quality evaluation index ZP are input into the feedback adjustment module, and the third wavelength s adjusted reflectance TF is output, and the subsequent processing of rice is optimized and the quality is controlled based on the influence of the third wavelength s adjusted reflectance TF on the spectral characteristics of the rice surface through the quality control decision module; The comprehensive comparison difference assessment module includes a rice surface spectrum difference unit under multiple wavelengths and a rice comprehensive quality assessment unit; The feedback adjustment module includes a rice surface spectrum analysis unit after feedback adjustment; The evaluation process of the rice comprehensive quality evaluation unit is as follows: ； in: ZP is the comprehensive quality assessment index; BC is the multi-wavelength spectral difference index; is the reflectivity at the second wavelength x; α is the logarithmic transformation adjustment coefficient, which is used to adjust slope, translation, and overall scaling of logarithmic transformations; ε is the avoidance coefficient, which is used to avoid negative infinity and zero value problems in logarithmic transformation; τ is the overall level adjustment coefficient, which acts on the degree of adjustment of the overall level of the comprehensive quality assessment index ZP; The multi-wavelength spectral difference index BC obtained by the rice surface spectral difference unit under multi-wavelength is input into the rice comprehensive quality assessment unit to obtain a comprehensive quality assessment index ZP that can identify the differences between different batches and varieties and conduct a comprehensive assessment of rice quality.

2. A quality detection method for rice processing according to claim 1, characterized in that: The difference manifestation process of the rice surface spectrum difference unit under the multi-wavelength is as follows: ； ； ； in: BC is the multi-wavelength spectral difference index; m is the total number of wavelengths; y is the first wavelength y, and y represents the multi-wavelength consideration interval from 1 to the total number of wavelengths m; is the reflectivity at the first wavelength y; BP is the average index of multi-wavelength reflectivity; FB is the multi-wavelength reflectivity standard deviation index; The reflectivity at the first wavelength y after processing by the data processing module , the multi-wavelength reflectivity average index BP and the multi-wavelength reflectivity standard deviation index FB are input into the rice surface spectrum difference unit under multi-wavelength to obtain the multi-wavelength spectrum difference index BC for evaluating the complexity and uniformity of rice surface characteristics.

3. A quality detection method for rice processing according to claim 2, characterized in that: The analysis process of the rice surface spectrum analysis unit after the feedback adjustment is as follows: ； in: TF is the reflectivity after adjustment at the third wavelength s; is the spectral characteristic index at the third wavelength s; δ is the adjustment factor, which reflects the influence of ZP on the reflectivity adjustment control; ZP is the comprehensive quality assessment index; MP is the comprehensive quality assessment target index; DB is the maximum index of comprehensive quality assessment, XB is the minimum index of comprehensive quality assessment, DB and XB are the maximum and minimum values ​​of the same batch and the same comprehensive quality assessment index ZP standard; The comprehensive quality evaluation index ZP obtained from the rice comprehensive quality evaluation unit and the reflectance after adjustment of the third wavelength s , comprehensive quality assessment target index MP, comprehensive quality assessment maximum index DB and comprehensive quality assessment minimum index XB are input into the rice surface spectral analysis unit after feedback adjustment to obtain the adjusted reflectance TF of the third wavelength s reflecting the influence of the comprehensive quality on the spectral characteristics.

4. A quality detection method for rice processing according to claim 3, characterized in that: Based on the adjusted reflectivity TF of the third wavelength s, the feedback adjustment analysis of the quality control decision module is as follows: S1, the reflectivity TF of the third wavelength s adjusted by the rice surface spectrum analysis unit after feedback adjustment is used as the reflectivity of the first wavelength y in the rice surface spectrum difference unit under the next round of multi-wavelengths Replace the data, calculate, and then output a new multi-wavelength spectral difference index ; S2, if the new multi-wavelength spectral difference index , compared with the value of the multi-wavelength spectral difference index BC, there is a more improved optimization trend, which means that the control optimization measures for rice processing quality are effective. In the subsequent processing, the same accuracy of processing optimization and quality control should be maintained; S3, if the new multi-wavelength spectral difference index , compared with the value of the multi-wavelength spectral difference index BC, there is a lower improvement trend, which means that the control and optimization measures for rice processing quality are effective, but not strong enough. In the subsequent processing, processing optimization and quality control should be strengthened; S4, if the new multi-wavelength spectral difference index , compared with the value of the multi-wavelength spectral difference index BC, there is no improvement trend, it is considered that the control optimization measures for rice processing quality are ineffective. In the subsequent processing process, the processing optimization strategy and quality control method are changed repeatedly until an effective improvement trend occurs.

5. A quality detection method for rice processing according to claim 1, characterized in that: The data preprocessing of the data processing module includes preprocessing operations of cleaning, denoising and normalizing the spectral data.

6. A quality detection method for rice processing according to claim 5, characterized in that: The equipment used in the spectrum data acquisition module includes a spectrometer, and the equipment used in the data processing module includes a computer, data processing software and quality control equipment.

Images