A smart monitoring system and method for cold fresh food safety storage

Abstract

The present invention discloses an intelligent monitoring system and method for the safe storage of fresh food, which specifically relates to the field of fresh food storage monitoring, and is used to solve the monitoring distortion and data inaccuracy problems caused by optical reflection interference. It is through accurate monitoring of light intensity and incident angle and dynamic adjustment of camera exposure parameters, combined with multi-band image acquisition and superposition analysis, it can effectively identify and locate optical reflection interference areas, and improve the adaptability and data accuracy of the monitoring system under complex light conditions. At the same time, by introducing the dynamic reflection stability index and the spectral scattering asymmetry coefficient, the authenticity of the reflection area is comprehensively evaluated, and areas where deceptive interference may exist are screened out, and the self-learning correction model is used to automatically adjust the image parameters to reduce the risk of reflection interference, thereby ensuring the clarity, authenticity and reliability of the monitoring picture, and significantly improving the monitoring accuracy and system stability of the fresh food storage environment.

Description

Technical Field

[0001] The present invention relates to the field of cold fresh food storage monitoring, and more specifically, to a cold fresh food safety storage intelligent monitoring system and method. Background Art

[0002] In the process of monitoring the cold and fresh food storage environment, the light conditions are complex and changeable, and the camera is often disturbed by the ambient light, resulting in overexposure, distortion or optical reflection interference in the image. These problems are particularly prominent in cold chain logistics and food safety management, because the clarity and accuracy of the monitoring image are directly related to the real-time monitoring and management effect of the storage environment, and any monitoring errors may lead to food safety hazards. Therefore, ensuring that high-quality monitoring images can be continuously obtained under complex light conditions is a key requirement of the cold and fresh food storage system.

[0003] However, traditional monitoring systems rely on manual adjustment of camera parameters or simple automatic exposure technology, which have obvious shortcomings when dealing with complex lighting environments. Manual adjustment is time-consuming and laborious and it is difficult to respond to light changes in real time. Traditional automatic exposure technology has limited effect in dealing with optical reflection interference and cannot effectively eliminate overexposure or distortion, resulting in reduced stability of the monitoring system and data accuracy. This technical deficiency is particularly prominent in the face of constantly changing storage environment light, so a more intelligent monitoring technology is urgently needed to solve the problem of optical reflection interference and improve system reliability and monitoring accuracy.

[0004] In order to solve the above problems, a technical solution is now provided. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide a smart monitoring system and method for the safe storage of fresh food. Through the precise monitoring of light intensity and incident angle, combined with the dynamic adjustment of camera exposure parameters, it can effectively identify and avoid overexposure or distortion caused by optical reflection, and significantly improve the clarity and stability of the monitoring picture under complex and changeable light conditions. Through multi-band image acquisition and superposition analysis, infrared, ultraviolet and other spectral data are combined with visible light images to accurately locate the optical reflection interference area, improve the system's recognition accuracy for hidden or weakened interference, and ensure the authenticity and integrity of the monitoring data. Further, through the comprehensive evaluation of the dynamic reflection stability index and the spectral scattering asymmetry coefficient, the reflection area with deceptive interference characteristics is effectively screened out to avoid misjudgment and ensure the safety of fresh food storage. Finally, through the application of the self-learning correction model, the system can automatically adjust the image parameters, reduce the impact of reflection interference on monitoring, reduce the complexity of manual intervention, improve the system's intelligent processing ability and long-term stability under complex light conditions, and comprehensively improve the monitoring accuracy and reliability of the fresh food storage environment to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] In a preferred embodiment, S1, monitoring light intensity and angle, adjusting exposure parameters of the camera, and identifying overexposed areas;

[0008] S2, enable the multi-band image acquisition function, superimpose infrared and ultraviolet spectrum data, compare and analyze with visible light images, and locate the optical reflection interference area;

[0009] S3, analyzing the fluctuation degree of the optical reflection interference area in the time series and whether the scattering characteristics under different spectra are symmetrical, and screening out the optical reflection interference area with deceptive interference;

[0010] S4, extracts the spectral features of the screened optical reflection interference area, uses historical data to train a self-learning correction model, and automatically adjusts image parameters.

[0011] In a preferred embodiment, step S1 includes the following contents:

[0012] S1.1, first, measure the light radiation intensity, which is calculated by combining the initial radiation brightness of the light source, the distance from the light source to the camera, and the reflection characteristics of the light source surface;

[0013] The monitoring of light radiation intensity and light incident angle is achieved by calculating the interaction between light radiation intensity and the light incident angle of the camera; first, the light radiation intensity is calculated ,in is the angle of incidence of the light, is the azimuth of the light source;

[0014] The light radiation intensity is calculated using the following formula: ;

[0015] in:

[0016] Indicates the initial radiance of the light source;

[0017] is the distance from the light source to the camera;

[0018] is the reflection coefficient;

[0019] S1.2, determining the camera exposure value according to the monitored light radiation intensity and light incident angle; to ensure that the image brightness is moderate, the camera exposure value is dynamically matched with the light radiation intensity and the light incident angle; when the camera exposure value is adjusted, the camera exposure value is controlled by the first and second adjustment coefficients so that it is in a predetermined proportional relationship with the light radiation intensity;

[0020] Camera exposure value Dynamically matching the light radiation intensity and the light incident angle, the camera exposure value adjustment formula is: ;

[0021] in:

[0022] and are the first and second adjustment factors, respectively;

[0023] S1.3, next, determine whether there is an overexposed area in the image; analyze the brightness of each pixel, and if its brightness value exceeds the image brightness threshold, it is identified as an overexposed area;

[0024] The identification of overexposed areas is based on the image brightness threshold For each pixel’s brightness , if the following conditions are met, the corresponding pixel is considered to be in the overexposed area: ;

[0025] in:

[0026] It is the third adjustment coefficient, which is used to balance the relationship between light radiation intensity and camera exposure value;

[0027] S1.4, after the calculation is completed, the light radiation intensity, camera exposure value and the recognition result of the overexposed area are used as the basis for subsequent processing.

[0028] In a preferred embodiment, step S2 includes the following contents:

[0029] S2.1, based on the light radiation intensity obtained in step S1 and camera exposure value , introducing the intensity coefficients of infrared and ultraviolet spectra and ,in and Represent the wavelengths of infrared and ultraviolet light respectively; the infrared and ultraviolet spectral intensity coefficients are constructed as follows:

[0030] ;

[0031] ;

[0032] in:

[0033] and is the intensity adjustment factor for infrared and ultraviolet spectra;

[0034] and is the scattering coefficient of infrared and ultraviolet light at different wavelengths;

[0035] Represents the overall adjustment factor of ambient light;

[0036] S2.2, analyzing the overexposed area identified in step S1 by superimposing multi-band spectra; superimposing the visible light image The intensity of the infrared and ultraviolet spectra is compared and analyzed: ;

[0037] in:

[0038] and is the weight coefficient when the infrared and ultraviolet spectra are superimposed;

[0039] is the composite image intensity, which is used to determine the reflection interference characteristics under different spectra;

[0040] S2.3, locate the optical reflection interference area by calculating the response intensity difference of the overexposed area in the multi-band image; if the response difference of the composite image intensity in the overexposed area is greater than or equal to the reflection interference threshold, the corresponding area is marked as the optical reflection interference area: ;

[0041] in:

[0042] is the maximum and minimum difference of the composite image intensity, which is used to determine the intensity of the reflection interference;

[0043] is the reflection interference threshold.

[0044] In a preferred embodiment, step S3 includes the following contents:

[0045] The generation logic of dynamic reflection stability index:

[0046] S3a.1, collecting spectral intensity data of the optical reflection interference area at multiple time points, and arranging the spectral intensity of each time point in sequence to form a time series;

[0047] S3a.2, calculate the spectral intensity difference between adjacent time points in the time series to obtain the intensity increment between each time point;

[0048] S3a.3, compare the spectral intensity increment with the intensity value at the previous time point and calculate the relative rate of change at each time point;

[0049] S3a.4, add the relative change rates of all time points in chronological order, accumulate the change amplitudes between all time points, and obtain the fluctuation accumulation value;

[0050] S3a.5, normalize the fluctuation accumulation value to obtain the final dynamic reflection stability index.

[0051] In a preferred embodiment, the generation logic of the spectral scattering asymmetry coefficient is:

[0052] S3b.1, first, calculate the scattering intensity of the optical reflection interference area in the infrared spectrum and the ultraviolet spectrum respectively; the scattering intensity of the infrared spectrum is obtained by multiplying the radiation intensity of the light source by the reflection coefficient in the area, and then combining it with a scattering function that depends on the wavelength of the spectrum; the scattering intensity of the ultraviolet spectrum is calculated using the same method;

[0053] S3b.2, next, the difference between the scattering intensities of the infrared and ultraviolet spectra is calculated; the scattering intensities of the two are compared to obtain the difference value between the two;

[0054] S3b.3, in order to quantify the asymmetry of the spectral scattering, the difference value of the scattering intensity is compared with the overall scattering intensity to obtain the intermediate value;

[0055] S3b.4, finally, the intermediate values ​​are normalized to obtain the spectral scattering asymmetry coefficient.

[0056] In a preferred embodiment, the reflection authenticity evaluation factor is obtained by combining the two indicators of dynamic reflection stability index and spectral scattering asymmetry coefficient and comprehensively calculating the two indicators.

[0057] In a preferred embodiment, step S4 includes the following contents:

[0058] S4.1, for the optical reflection interference area whose reflection authenticity evaluation factor is less than the reflection authenticity threshold, the spectral characteristics of the optical reflection interference area in multiple bands are extracted by hyperspectral imaging technology;

[0059] S4.2, collect the monitoring data accumulated by the monitoring system in the past period of time, and use data mining technology to clean and organize these historical data to remove noise and irrelevant information;

[0060] S4.3, using a self-learning algorithm based on machine learning, combined with the extracted spectral features and processed historical data, to train a self-learning correction model; the training process of the model is completed through supervised learning, where the input is the screened spectral features and the output is the corresponding image adjustment parameters;

[0061] S4.4, after the model training is completed, the model is calibrated and optimized;

[0062] S4.5, deploys a trained and optimized self-learning correction model and integrates it into the monitoring system; when an optical reflection interference area is detected, the image parameters of the camera are automatically adjusted.

[0063] A smart monitoring system for cold fresh food safety storage, comprising: a light monitoring module, a multi-frequency imaging module, a fluctuation analysis module and a correction model module;

[0064] Light monitoring module: monitors light intensity and angle, adjusts camera exposure parameters, and identifies overexposed areas; the identified overexposed area data is transmitted to the multi-band imaging module;

[0065] Multi-band imaging module: Enable the multi-band imaging acquisition function, superimpose infrared and ultraviolet spectrum data, compare and analyze with visible light images, and locate the optical reflection interference area; the located optical reflection interference area data is transmitted to the fluctuation analysis module;

[0066] Fluctuation analysis module: Analyze the fluctuation degree of the optical reflection interference area in the time series and whether the scattering characteristics under different spectra are symmetrical, and screen out the optical reflection interference area with deceptive interference; the screened interference area data is transmitted to the correction model module;

[0067] Correction model module: extracts the spectral characteristics of the screened optical reflection interference area, uses historical data to train the self-learning correction model, and automatically adjusts the image parameters; the adjusted image parameters are fed back to the light monitoring module for cyclic optimization.

[0068] The technical effects and advantages of the intelligent monitoring system and method for cold fresh food safety storage of the present invention are as follows:

[0069] 1. The present invention can effectively identify and avoid overexposure or distortion areas in the picture by accurately monitoring the light intensity and incident angle, combined with a method of dynamically adjusting the camera exposure parameters. It significantly improves the clarity and detail retention of the monitoring picture, especially in an environment with complex and changeable light, ensures the stability and reliability of the monitoring image, and reduces the risk of misjudgment caused by optical reflection interference.

[0070] 2. The present invention combines infrared, ultraviolet and other spectral data with visible light images by introducing multi-band image acquisition and superposition analysis, and can accurately locate specific areas of optical reflection interference. This not only improves the monitoring system's adaptability to complex optical conditions, but also enhances the recognition accuracy of hidden or weakened interference, thereby ensuring the authenticity and integrity of the monitoring images, and providing a higher guarantee for the effective monitoring of cold and fresh food safety storage.

[0071] 3. The present invention can effectively identify and screen out deceptive interference in the optical reflection interference area through step S3, ensuring that the monitoring system can still provide true and reliable monitoring data under complex light conditions. By using the comprehensive evaluation of the dynamic reflection stability index and the spectral scattering asymmetry coefficient, the authenticity of the reflection area can be accurately judged, avoiding misjudgment and safety hazards caused by abnormal or unstable optical reflection, thereby improving the monitoring accuracy and safety of the cold fresh food storage environment.

[0072] 4. The present invention can automatically identify and adjust the image parameters of the optical reflection interference area through step S4, effectively reducing the impact of reflection interference on the monitoring screen. The application of the self-learning correction model not only improves the clarity of the monitoring screen and the accuracy of the data, but also reduces the complexity of manual adjustment and the risk of misjudgment. This method enables the monitoring system to have stronger adaptability and intelligent processing capabilities when facing complex light conditions, and improves the overall stability and reliability of the cold fresh food storage environment monitoring system. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 A schematic diagram of a flow chart of a method for intelligent monitoring of cold fresh food storage safety according to the present invention;

[0074] Figure 2 This is a structural schematic diagram of a cold fresh food safety storage intelligent monitoring system of the present invention. DETAILED DESCRIPTION

[0075] 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.

[0076] Example 1

[0077] Figure 1 The present invention provides a method for intelligent monitoring of cold fresh food storage safety, comprising:

[0078] S1, monitors light intensity and angle, adjusts camera exposure parameters, and identifies overexposed areas;

[0079] S2, enable the multi-band image acquisition function, superimpose infrared and ultraviolet spectrum data, compare and analyze with visible light images, and locate the optical reflection interference area;

[0080] S3, analyzing the fluctuation degree of the optical reflection interference area in the time series and whether the scattering characteristics under different spectra are symmetrical, and screening out the optical reflection interference area with deceptive interference;

[0081] S4, extracts the spectral features of the screened optical reflection interference area, uses historical data to train a self-learning correction model, and automatically adjusts image parameters.

[0082] In the cold food storage environment, due to the complex and changeable light conditions, especially when the light intensity and angle are unstable, the surveillance camera is easily disturbed by optical reflection, resulting in overexposure or distortion of the image. In order to ensure the clarity and accuracy of the surveillance image, it is necessary to accurately monitor the ambient light intensity and incident angle, and dynamically adjust the camera's exposure parameters to achieve real-time control of the image quality. This step is designed to reduce the impact of optical reflection on the surveillance image, ensuring that a stable and clear surveillance image can be provided under different light conditions.

[0083] Step S1 includes the following contents:

[0084] S1.1, first, measure the light radiation intensity, considering the impact of the light incident angle and the azimuth of the light source on the intensity. The calculation of the light radiation intensity combines the initial radiation brightness of the light source, the distance from the light source to the camera, and the reflective characteristics of the light source surface. The light intensity will weaken with the increase of distance. At the same time, the difference in the incident angle and azimuth will also affect the distribution of the reflection intensity.

[0085] The monitoring of light radiation intensity and light incident angle is achieved by calculating the interaction between light radiation intensity and the light incident angle of the camera. First, calculate the light radiation intensity ,in is the angle of incidence of the light, is the azimuth of the light source.

[0086] The light radiation intensity is calculated using the following formula: ;

[0087] in:

[0088] Represents the initial radiance of the light source.

[0089] is the distance from the light source to the camera.

[0090] is the reflection coefficient, which represents the reflection characteristics of the light source surface.

[0091] S1.2, determine the camera exposure value based on the monitored light radiation intensity and light incident angle. To ensure that the image brightness is moderate, the camera exposure value needs to be dynamically matched with the light radiation intensity and light incident angle. When adjusting the camera exposure value, it is necessary to consider changes in environmental conditions and control the camera exposure value by adjusting the coefficient so that it is in a predetermined proportional relationship with the light radiation intensity. This can prevent overexposure or distortion of the image due to excessive light.

[0092] In order to avoid overexposure or distortion of the image, the camera exposure value It should be dynamically matched with the light radiation intensity and light incident angle. The camera exposure value adjustment formula is: ;

[0093] in:

[0094] and They are the first and second adjustment coefficients, which are adjusted according to different environmental conditions to ensure that the image captured by the camera has moderate brightness.

[0095] S1.3, next, determine whether there is an overexposed area in the image. Analyze the brightness of each pixel. If its brightness value exceeds the image brightness threshold and is greatly affected by the light intensity, it is considered an overexposed area. This method can accurately locate areas with abnormal image brightness due to optical reflection or excessive light intensity, avoiding loss of picture details.

[0096] The identification of overexposed areas is based on the image brightness threshold For each pixel's brightness , if the following conditions are met, the pixel is considered to be in the overexposed area: ;

[0097] in:

[0098] It is the third adjustment coefficient, which is used to balance the relationship between light radiation intensity and camera exposure value.

[0099] S1.4, after the calculation is completed, the light radiation intensity, camera exposure value and the identification results of the overexposed area are used as the basis for subsequent processing. These values ​​are used to further optimize the monitoring image quality, reduce the negative impact of optical reflection interference on the monitoring effect, and ensure that the image clarity and accuracy can be maintained under complex lighting conditions.

[0100] The present invention can effectively identify and avoid overexposed or distorted areas in the image by accurately monitoring the light intensity and incident angle, combined with a method for dynamically adjusting the camera exposure parameters. It significantly improves the clarity and detail retention of the monitoring image, especially in an environment with complex and changeable light, ensures the stability and reliability of the monitoring image, and reduces the risk of misjudgment caused by optical reflection interference.

[0101] In the process of monitoring the cold food storage environment, a single spectrum visible light image is often difficult to fully capture the optical reflection interference in the environment, especially under the influence of different light wavelengths on camera imaging, some interference may be hidden or weakened. In order to more accurately identify and locate these optical reflection interference areas, it is necessary to combine infrared, ultraviolet and other multi-band spectral data, and reveal the interference characteristics under different spectra through the superposition and comparative analysis of multi-band images. The design of this step aims to utilize the complementarity of multi-spectral data to improve the sensitivity and positioning accuracy of optical reflection interference, and ensure that the monitoring system can still maintain a high degree of image clarity and authenticity under complex optical conditions.

[0102] Step S2 includes the following contents:

[0103] By using the light radiation intensity, camera exposure value and overexposed area recognition results calculated in step S1, multi-band image analysis is further performed to accurately locate the specific area of ​​optical reflection interference. Multi-band image analysis combines infrared, ultraviolet and other spectral data, and superimposes and compares them with visible light images to ensure that interference features under different spectra can be identified, ultimately achieving accurate positioning of the reflection interference area.

[0104] S2.1, based on the light radiation intensity obtained in step S1 and camera exposure value , introducing the intensity coefficients of infrared and ultraviolet spectra and ,in and Represent the wavelengths of infrared and ultraviolet light respectively. The infrared and ultraviolet spectral intensity coefficients are constructed as follows:

[0105] ;

[0106] ;

[0107] in:

[0108] and is the intensity adjustment factor for infrared and ultraviolet spectra.

[0109] and is the scattering coefficient of infrared and ultraviolet light at different wavelengths.

[0110] Represents the overall adjustment factor of ambient light, which is fine-tuned according to actual environmental conditions.

[0111] S2.2, analyzing the overexposed area identified in step S1 by superimposing multiple frequency band spectra. Superimposing visible light images The intensity of the infrared and ultraviolet spectra is compared and analyzed: ;

[0112] in:

[0113] and is the weight coefficient when the infrared and ultraviolet spectra are superimposed.

[0114] It is the composite image intensity, which is used to determine the reflection interference characteristics under different spectra.

[0115] S2.3, locate the optical reflection interference area by calculating the response intensity difference of the overexposed area in the multi-band image. If the response difference of the composite image intensity in the overexposed area is greater than or equal to the reflection interference threshold, the area is marked as the optical reflection interference area: ;

[0116] in:

[0117] It is the maximum and minimum difference of the composite image intensity, which is used to judge the intensity of reflection interference.

[0118] is the reflection interference threshold, which determines whether there is significant optical reflection interference.

[0119] The present invention combines infrared, ultraviolet and other spectral data with visible light images by introducing multi-band image acquisition and superposition analysis, and can accurately locate specific areas of optical reflection interference. This not only improves the monitoring system's adaptability to complex optical conditions, but also enhances the recognition accuracy of hidden or weakened interference, thereby ensuring the authenticity and integrity of the monitoring image, and providing a higher guarantee for the effective monitoring of cold and fresh food safety storage.

[0120] In the process of monitoring the storage environment of cold fresh food, it is crucial to judge the authenticity of the optical reflection interference area, especially under complex light conditions, the reflection area may fluctuate and be asymmetric, and these abnormal characteristics may conceal the true storage conditions or cause misjudgment. In order to accurately evaluate the reflection authenticity of these areas, it is necessary to combine the dynamic reflection stability index in the time dimension and the scattering asymmetry coefficient in the spectral dimension, and through the comprehensive calculation of the reflection authenticity evaluation factor, further screen and process the optical reflection interference areas that may have deceptive interference, to ensure the reliability of the monitoring system and the credibility of the data.

[0121] Step S3 includes the following contents:

[0122] During the monitoring of cold food storage environments, the reflection intensity of the optical reflection interference area may fluctuate significantly over time. This volatility may not only affect the stability of monitoring, but also cover up or amplify certain key details, leading to misjudgment or neglect of potential safety hazards. The dynamic reflection stability index can accurately evaluate the degree of fluctuation of the reflection area in the time series, provide a basis for identifying possible deceptive interference, and ensure the reliability of the monitoring system under complex lighting conditions.

[0123] The generation logic of dynamic reflection stability index:

[0124] S3a.1, collect the spectral intensity data of the optical reflection interference area at multiple time points, and arrange the spectral intensity at each time point in sequence to form a time series. The construction of the time series is to analyze the change trend of the optical reflection interference area in the time dimension for subsequent incremental calculation.

[0125] S3a.2, calculate the spectral intensity difference between adjacent time points in the time series to obtain the intensity increment between each time point. The spectral intensity increment reflects the change amplitude of the optical reflection interference area between two consecutive time points and is the basis for judging the reflection stability of the area.

[0126] S3a.3, compare the spectral intensity increment with the intensity value at the previous time point, and calculate the relative change rate at each time point. The relative change rate is used to quantify the degree of change in spectral intensity, further reflecting the volatility of the region in the time series.

[0127] S3a.4, add the relative change rates of all time points in chronological order, accumulate the change amplitudes between all time points, and obtain the fluctuation accumulation value. Through integration, the fluctuation accumulation value of the optical reflection interference area during the entire monitoring period can be obtained. This value is used to determine whether the area has a large volatility, thereby indicating potential deceptive interference.

[0128] S3a.5, normalize the fluctuation accumulation value to obtain the final dynamic reflection stability index. The normalized index value can intuitively reflect the stability of the optical reflection interference area. The higher the value, the more unstable the reflection change in the area, and there may be deceptive interference.

[0129] The dynamic reflection stability index is used to indicate and reflect the stability of the spectral intensity change of the optical reflection interference area during the entire monitoring period. In the process of screening and analyzing whether the optical reflection interference area obtained in step S2 has deceptive properties, the larger the index, the more unstable the spectral intensity change in the area, the larger the fluctuation range, which may indicate that there is deceptive interference in the area; conversely, the smaller the index, the more stable the spectral intensity change, the smaller the fluctuation range, indicating that the optical reflection interference in the area is more likely to be a real reflection phenomenon rather than a pseudo-interference caused by other factors. Through the analysis of the dynamic reflection stability index, it is possible to effectively judge and screen out optical reflection interference areas that may have deceptive properties.

[0130] In optical monitoring, the spectral scattering characteristics of the optical reflection interference area are the key factors in judging its authenticity and consistency. The reflection asymmetry under different spectra (such as infrared and ultraviolet) may indicate that there is an abnormality in the area, or interference caused by special optical effects in the environment. The introduction of the spectral scattering asymmetry coefficient can quantify the degree of asymmetry, thereby helping to identify and locate potential deceptive reflection interference, ensuring that the optical reflection interference area in the monitoring image is real and reliable, and avoiding safety hazards caused by misjudgment.

[0131] The generation logic of the spectral scattering asymmetry coefficient is:

[0132] S3b.1, first, the scattering intensity of the optical reflection interference area under the infrared spectrum and ultraviolet spectrum needs to be calculated respectively. The scattering intensity of the infrared spectrum is obtained by multiplying the radiation intensity of the light source with the reflection coefficient in the area, and then combining it with a scattering function that depends on the wavelength of the spectrum. The scattering intensity of the ultraviolet spectrum is calculated in the same way. Through this calculation, the scattering characteristics of the area under the infrared and ultraviolet spectra can be obtained.

[0133] S3b.2, next, calculate the difference between the scattering intensities of the infrared and ultraviolet spectra. Compare the scattering intensities of the two to obtain the difference between them. This difference reflects whether the response of the optical reflection interference area under different spectra is consistent. The greater the difference, the more asymmetric the scattering characteristics of the two spectra.

[0134] S3b.3, in order to quantify the asymmetry of spectral scattering, the difference in scattering intensity is compared with the overall scattering intensity to obtain an intermediate value. This coefficient is used to measure whether the scattering characteristics of infrared and ultraviolet spectra in the same area are symmetrical. The higher the value, the stronger the asymmetry of the scattering characteristics.

[0135] S3b.4, finally, the intermediate values ​​are normalized to obtain the spectral scattering asymmetry coefficient to ensure a consistent comparison standard under different spectral combinations. The normalized coefficient is used to determine whether there are abnormal reflection characteristics in the optical reflection interference area. Areas with higher values ​​may have deceptive interference and therefore require special attention.

[0136] The spectral scattering asymmetry coefficient is used to indicate and reflect whether the scattering characteristics of the optical reflection interference area under different spectra are symmetrical. The larger the coefficient, the greater the asymmetry between the scattering characteristics of the infrared and ultraviolet spectra in this area, which may indicate that the optical reflection characteristics of this area are abnormal or there is deceptive interference; the smaller the coefficient, the more symmetrical the scattering characteristics, indicating that the optical reflection of this area is more stable and consistent, and it may belong to the normal reflection area.

[0137] By combining the dynamic reflection stability index (DRSI) and the spectral scattering asymmetry coefficient (ASI), the reflection authenticity evaluation factor is obtained through comprehensive calculation of the two indicators. The authenticity of the optical reflection interference area is evaluated. The evaluation process adopts a weighted summation method to comprehensively reflect the impact of these two indicators on the reflection authenticity.

[0138] For example, it can be calculated by the following formula: ;

[0139] is the reflection authenticity assessment factor, and are the dynamic reflex stability index and the dynamic reflex stability index, and They are respectively the dynamic reflection stability index and the preset proportional coefficient of the dynamic reflection stability index, and both are greater than zero.

[0140] The reflection authenticity assessment factor is used to represent and reflect the reflection authenticity and credibility of the optical reflection interference area. By combining the two indicators of dynamic reflection stability index and spectral scattering asymmetry coefficient, it comprehensively evaluates whether the optical reflection of the area is abnormal or deceptive. The larger the factor value, the lower the reflection authenticity of the area, the possible deceptive interference, and the higher the instability and asymmetry of the reflection characteristics; the smaller the factor value, the more authentic and reliable the optical reflection of the area is, the stable and symmetrical reflection characteristics are, and the higher the credibility is.

[0141] When comparing the reflection authenticity assessment factor with the reflection authenticity classification threshold, if the reflection authenticity assessment factor is greater than or equal to the reflection authenticity threshold, it means that the authenticity of the optical reflection interference area is low, and there is a high possibility of deceptive interference. The reflection characteristics may be abnormal or unstable and need further review and processing. On the contrary, if the reflection authenticity assessment factor is less than the reflection authenticity threshold, it means that the authenticity of the optical reflection interference area is high, the reflection characteristics are stable and credible, and it is a normal optical reflection, and the monitoring data can be trusted.

[0142] The present invention can effectively identify and screen out deceptive interference in the optical reflection interference area through step S3, ensuring that the monitoring system can still provide true and reliable monitoring data under complex light conditions. By using the comprehensive evaluation of the dynamic reflection stability index and the spectral scattering asymmetry coefficient, the authenticity of the reflection area can be accurately judged, avoiding misjudgment and safety hazards caused by abnormal or unstable optical reflection, thereby improving the monitoring accuracy and safety of the cold fresh food storage environment.

[0143] In the process of monitoring the cold food storage environment, optical reflection interference often leads to a decline in the quality of the monitoring image, affecting the accuracy and reliability of the data. The traditional method of manually adjusting image parameters cannot cope with the complex and changeable optical reflection conditions. Therefore, it is necessary to introduce a self-learning correction model based on spectral characteristics to achieve intelligent processing of optical reflection interference through training with historical data and real-time adjustment. This step aims to ensure that the monitoring system can still provide clear and accurate images in complex lighting environments through automated image parameter optimization, avoiding monitoring distortion caused by reflection interference.

[0144] Step S4 includes the following contents:

[0145] S4.1, for the optical reflection interference area where the reflection authenticity assessment factor is less than the reflection authenticity threshold, the spectral data of the corresponding optical reflection interference area is deeply analyzed. Through hyperspectral imaging technology, the spectral characteristics of the optical reflection interference area in multiple bands are extracted, including detailed information such as spectral intensity, wavelength distribution, and scattering pattern. These features will be used for subsequent model training to help the system identify and handle similar reflection interference situations.

[0146] S4.2, collect monitoring data accumulated by the monitoring system over the past period of time, especially historical records related to optical reflection interference areas. Through data mining technology, these historical data are cleaned and sorted to remove noise and irrelevant information to ensure that the data set for model training is accurate and representative. The collection and processing of historical data is to enable the self-learning model to adapt to various reflection interference scenarios and improve the generalization ability of the model.

[0147] S4.3, a self-learning correction model is trained by combining the extracted spectral features and processed historical data using a self-learning algorithm based on machine learning, such as convolutional neural network (CNN) or long short-term memory network (LSTM) in deep learning. The training process of the model is completed through supervised learning, where the input is the screened spectral features and the output is the corresponding image adjustment parameters. The purpose of training is to enable the model to automatically identify the optical reflection interference area and generate the best image parameter adjustment solution.

[0148] S4.4, after the model training is completed, the model is corrected and optimized. Through cross-validation and algorithm tuning, the accuracy and response speed of the model are further improved to ensure that the optical reflection interference area can be accurately and quickly identified and processed in real-time monitoring. During the optimization process, an adaptive learning mechanism can also be introduced to enable the model to continuously adjust and optimize itself as the monitoring environment changes.

[0149] S4.5, deploys a trained and optimized self-learning correction model and integrates it into the monitoring system. When the system detects an area of ​​optical reflection interference, the model will be automatically called to automatically adjust the camera's image parameters, such as exposure, contrast, color temperature, etc., based on real-time spectral feature data and experience learned during training, to reduce or eliminate the impact of reflection interference on the monitoring image. Through this automatic adjustment mechanism, the quality and clarity of the monitoring image can be significantly improved, ensuring that accurate monitoring data can still be obtained under complex lighting conditions.

[0150] S4.6, in actual application, the self-learning correction model will continue to learn new monitoring data, and regularly update and iterate the model. By introducing a continuous learning mechanism, it ensures that the model can adapt to new optical reflection interference scenarios, and continuously improves the ability to identify and process reflection interference, maintaining the long-term stability and efficiency of the monitoring system.

[0151] Through step S4, the present invention can automatically identify and adjust the image parameters of the optical reflection interference area, effectively reducing the impact of reflection interference on the monitoring screen. The application of the self-learning correction model not only improves the clarity of the monitoring screen and the accuracy of the data, but also reduces the complexity of manual adjustment and the risk of misjudgment. This method enables the monitoring system to have stronger adaptability and intelligent processing capabilities when facing complex light conditions, and improves the overall stability and reliability of the cold fresh food storage environment monitoring system.

[0152] Figure 2 The present invention provides a cold fresh food storage safety intelligent monitoring system, comprising: a light monitoring module, a multi-frequency imaging module, a fluctuation analysis module and a correction model module;

[0153] Light monitoring module: monitors light intensity and angle, adjusts camera exposure parameters, and identifies overexposed areas; the identified overexposed area data is transmitted to the multi-band imaging module;

[0154] Multi-band imaging module: Enable the multi-band imaging acquisition function, superimpose infrared and ultraviolet spectrum data, compare and analyze with visible light images, and locate the optical reflection interference area; the located optical reflection interference area data is transmitted to the fluctuation analysis module;

[0155] Fluctuation analysis module: Analyze the fluctuation degree of the optical reflection interference area in the time series and whether the scattering characteristics under different spectra are symmetrical, and screen out the optical reflection interference area with deceptive interference; the screened interference area data is transmitted to the correction model module;

[0156] Correction model module: extracts the spectral characteristics of the screened optical reflection interference area, uses historical data to train the self-learning correction model, and automatically adjusts the image parameters; the adjusted image parameters are fed back to the light monitoring module for cyclic optimization.

[0157] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0158] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0159] It should be noted that, in this article, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "including a..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0160] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for intelligent monitoring of cold fresh food storage safety, characterized in that: Includes steps: S1, monitors light intensity and angle, adjusts camera exposure parameters, and identifies overexposed areas; S2, enable the multi-band image acquisition function, superimpose infrared and ultraviolet spectrum data, compare and analyze with visible light images, and locate the optical reflection interference area; S3, analyzing the fluctuation degree of the optical reflection interference area in the time series and whether the scattering characteristics under different spectra are symmetrical, and screening out the optical reflection interference area with deceptive interference; S4, extracting the spectral features of the screened optical reflection interference area, using historical data to train a self-learning correction model, and automatically adjusting image parameters; Step S1 includes the following contents: S1.1, first, measure the light radiation intensity, which is calculated by combining the initial radiation brightness of the light source, the distance from the light source to the camera, and the reflection characteristics of the light source surface; The monitoring of light radiation intensity and light incident angle is achieved by calculating the interaction between light radiation intensity and the light incident angle of the camera; first, the light radiation intensity is calculated ,in is the angle of incidence of the light, is the azimuth of the light source; The light radiation intensity is calculated using the following formula: ; in: Indicates the initial radiance of the light source; is the distance from the light source to the camera; is the reflection coefficient; S1.2, determining the camera exposure value according to the monitored light radiation intensity and light incident angle; to ensure that the image brightness is moderate, the camera exposure value is dynamically matched with the light radiation intensity and the light incident angle; when the camera exposure value is adjusted, the camera exposure value is controlled by the first and second adjustment coefficients so that it is in a predetermined proportional relationship with the light radiation intensity; Camera exposure value Dynamically matching the light radiation intensity and the light incident angle, the camera exposure value adjustment formula is: ; in: and are the first and second adjustment factors, respectively; S1.3, next, determine whether there is an overexposed area in the image; analyze the brightness of each pixel, and if its brightness value exceeds the image brightness threshold, it is identified as an overexposed area; The identification of overexposed areas is based on the image brightness threshold For each pixel’s brightness , if the following conditions are met, the corresponding pixel is considered to be in the overexposed area: ； in: It is the third adjustment coefficient, which is used to balance the relationship between light radiation intensity and camera exposure value; S1.4, after the calculation is completed, the light radiation intensity, camera exposure value and the identification result of the overexposed area are used as the basis for subsequent processing; Step S2 includes the following contents: S2.1, based on the light radiation intensity obtained in step S1 and camera exposure value , introducing the intensity coefficients of infrared and ultraviolet spectra and ,in and Represent the wavelengths of infrared and ultraviolet light respectively; the infrared and ultraviolet spectral intensity coefficients are constructed as follows: ； ； in: and is the intensity adjustment factor for infrared and ultraviolet spectra; and is the scattering coefficient of infrared and ultraviolet light at different wavelengths; Represents the overall adjustment factor of ambient light; S2.2, analyzing the overexposed area identified in step S1 by superimposing multi-band spectra; superimposing the visible light image The intensity of the infrared and ultraviolet spectra is compared and analyzed: ； in: and is the weight coefficient when the infrared and ultraviolet spectra are superimposed; is the composite image intensity, which is used to determine the reflection interference characteristics under different spectra; S2.3, locate the optical reflection interference area by calculating the response intensity difference of the overexposed area in the multi-band image; if the response difference of the composite image intensity in the overexposed area is greater than or equal to the reflection interference threshold, the corresponding area is marked as the optical reflection interference area: ; in: is the maximum and minimum difference of the composite image intensity, which is used to determine the intensity of the reflection interference; is the reflection interference threshold.

2. The method for intelligent monitoring of cold fresh food safety storage according to claim 1, characterized in that: Step S3 includes the following contents: The generation logic of dynamic reflection stability index: S3a.1, collecting spectral intensity data of the optical reflection interference area at multiple time points, and arranging the spectral intensity of each time point in sequence to form a time series; S3a.2, calculate the spectral intensity difference between adjacent time points in the time series to obtain the intensity increment between each time point; S3a.3, compare the spectral intensity increment with the intensity value at the previous time point and calculate the relative rate of change at each time point; S3a.4, add the relative change rates of all time points in chronological order, accumulate the change amplitudes between all time points, and obtain the fluctuation accumulation value; S3a.5, normalize the fluctuation accumulation value to obtain the final dynamic reflection stability index.

3. The method for intelligent monitoring of cold fresh food safety storage according to claim 2, characterized in that: The generation logic of the spectral scattering asymmetry coefficient is: S3b.1, first, calculate the scattering intensity of the optical reflection interference area in the infrared spectrum and the ultraviolet spectrum respectively; the scattering intensity of the infrared spectrum is obtained by multiplying the radiation intensity of the light source by the reflection coefficient in the area, and then combining it with a scattering function that depends on the wavelength of the spectrum; the scattering intensity of the ultraviolet spectrum is calculated using the same method; S3b.2, next, the difference between the scattering intensities of the infrared and ultraviolet spectra is calculated; Compare the scattering intensities of the two and obtain the difference between them; S3b.3, in order to quantify the asymmetry of the spectral scattering, the difference value of the scattering intensity is compared with the overall scattering intensity to obtain the intermediate value; S3b.4, finally, the intermediate values ​​are normalized to obtain the spectral scattering asymmetry coefficient.

4. The method for intelligent monitoring of cold fresh food safety storage according to claim 3, characterized in that: By combining the dynamic reflection stability index and the spectral scattering asymmetry coefficient, the reflection authenticity evaluation factor is obtained through comprehensive calculation of the two indicators.

5. The method for intelligent monitoring of cold fresh food safety storage according to claim 4, characterized in that: Step S4 includes the following contents: S4.1, for the optical reflection interference area whose reflection authenticity evaluation factor is less than the reflection authenticity threshold, the spectral characteristics of the optical reflection interference area in multiple bands are extracted by hyperspectral imaging technology; S4.2, collect the monitoring data accumulated by the monitoring system in the past period of time, and use data mining technology to clean and organize these historical data to remove noise and irrelevant information; S4.3, using a self-learning algorithm based on machine learning, combined with the extracted spectral features and processed historical data, to train a self-learning correction model; the training process of the model is completed through supervised learning, where the input is the screened spectral features and the output is the corresponding image adjustment parameters; S4.4, after the model training is completed, the model is calibrated and optimized; S4.5, deploys a trained and optimized self-learning correction model and integrates it into the monitoring system; when an optical reflection interference area is detected, the image parameters of the camera are automatically adjusted.

6. A smart monitoring system for the safe storage of fresh food, used to implement the smart monitoring method for the safe storage of fresh food according to any one of claims 1 to 5, characterized in that: include: Light monitoring module, multi-frequency imaging module, fluctuation analysis module and correction model module; Light monitoring module: monitors light intensity and angle, adjusts camera exposure parameters, and identifies overexposed areas; the identified overexposed area data is transmitted to the multi-band imaging module; Multi-band imaging module: Enable the multi-band imaging acquisition function, superimpose infrared and ultraviolet spectrum data, compare and analyze with visible light images, and locate the optical reflection interference area; the located optical reflection interference area data is transmitted to the fluctuation analysis module; Fluctuation analysis module: Analyze the fluctuation degree of the optical reflection interference area in the time series and whether the scattering characteristics under different spectra are symmetrical, and screen out the optical reflection interference area with deceptive interference; the screened interference area data is transmitted to the correction model module; Correction model module: extracts the spectral characteristics of the screened optical reflection interference area, uses historical data to train the self-learning correction model, and automatically adjusts the image parameters; the adjusted image parameters are fed back to the light monitoring module for cyclic optimization.

Images