Method and system for detecting COD (Chemical Oxygen Demand) parameters of water quality by using laser-induced fluorescence imaging method

Abstract

The invention belongs to the technical field of chemical oxygen demand detection, and discloses a method and a system for detecting water quality COD (Chemical Oxygen Demand) parameters by a laser-induced fluorescence imaging method, and the method comprises the following steps: mixing sodium humate and glucose according to a specific ratio to prepare a detection solution for calibration; the method comprises the following steps: exciting a detection solution by using a laser-induced fluorescence technology, acquiring a fluorescence image by using an image sensor, and carrying out feature extraction and processing on the fluorescence image by combining an image processing technology; the RGB color characteristic value of the selected fluorescence area is analyzed, and after a COD concentration prediction model based on the RGB color characteristic average value of the fluorescence image characteristic area is constructed through a single dependent variable partial least square regression analysis method, COD water quality parameters of an actual water sample are detected. The method is suitable for detecting low-concentration organic matters in water, has the advantages of high sensitivity, high stability, non-contact property, high detection speed and the like, and has a good application value in the aspect of rapid water quality monitoring and early warning.

Description

Technical Field

[0001] The present invention belongs to the technical field of chemical oxygen demand detection, and particularly relates to a method and system for detecting water quality COD parameters by laser-induced fluorescence imaging method. Background Technique

[0002] In the detection and monitoring of water quality, chemical oxygen demand (COD) is an important indicator in water quality monitoring, which reflects the degree of organic pollution in water bodies. The existing standard method for detecting water quality COD is still the chemical method. For example, in a Chinese patent, the authorization announcement number is: CN 112730543 B, and the authorization announcement date is: January 17, 2023. It discloses a method for constructing a portable potentiometric photoelectrochemical sensor for rapid detection of chemical oxygen demand. By preparing CuS / TiO 2 As a photoactive material, a potentiometric photoelectrochemical sensor was established to realize the analysis and detection of chemical oxygen demand. Although the above chemical detection method has high accuracy, it has defects such as high sample procurement cost, cumbersome operation, long measurement cycle, and easy to cause secondary pollution, and it is difficult to meet the requirements of daily real-time detection and analysis of water quality. Compared with the chemical method, the optical method has become a hot spot in water quality detection due to its advantages of high detection efficiency, high precision, pollution-free, and convenience.

[0003] Currently, the optical methods mainly include absorption spectroscopy, hyperspectral analysis, and fluorescence spectroscopy. The essence of the absorption spectroscopy method is to calculate the COD value of water by measuring the absorption rate of organic matter at a certain wavelength, which is a common method for online measuring COD; however, when the concentration of organic matter in the solution is low, the measurement accuracy of this method is poor, the sensitivity is low, and the detection accuracy for lower concentrations is relatively poor. The hyperspectral analysis method has the advantages of high spatial resolution, high spectral resolution, and spectral integration, but this method has complex technical operations and high costs, and is currently mainly used for qualitative detection and analysis of water quality. The detection sensitivity of the fluorescence spectroscopy method is higher than that of the absorption spectroscopy method, and the detection and analysis speed is also higher than that of the absorption photometry method; however, compared with the laboratory standard solution, in the monitoring process, the types of organic matter are complex, and the differences in the fluorescence peaks and band ranges of each organic matter will also lead to the appearance of fluorescence peak overlap phenomenon. Therefore, there are certain errors in the results of simple fluorescence spectroscopy analysis.

[0004] Through the above analysis, the problems and defects existing in the prior art are as follows:

[0005] (1) The traditional COD chemical detection method has defects such as cumbersome operation, long measurement cycle, and easy to cause secondary pollution, and at the same time, it is difficult to meet the requirements of daily real-time detection and analysis of water quality.

[0006] (2) In existing optical methods, when the concentration of organic matter in the solution is low, the absorption spectrometry has poor measurement accuracy and low sensitivity; the hyperspectral analysis method is complex in operation and high in cost, and is currently mainly used for qualitative detection and analysis of water quality; while the simple fluorescence spectrometry has certain errors in results.

[0007] (3) In existing water quality detection systems applying image processing technology, absorption spectrometry or reflection spectrometry is mostly adopted, while the combination of fluorescence spectrometry and image processing is relatively less applied, and the detection error for low-concentration organic matter is relatively large.

[0008] The main advantages of fluorescence analysis method are high sensitivity and non-contact detection, while image processing technology not only can detect non-contact but also has advantages such as large field of view and low cost. Therefore, there is an urgent need for a method that combines fluorescence analysis method and image processing technology to detect the COD parameter of water quality. Summary of the Invention

[0009] To overcome the problems existing in the related technologies, the disclosed embodiments of the present invention provide a method and system for detecting COD parameter of water quality by laser-induced fluorescence imaging method, and the technical solutions are as follows:

[0010] The present invention is implemented as follows. A method for detecting COD parameter of water quality by laser-induced fluorescence imaging method, which combines fluorescence analysis method and image processing technology to detect the COD concentration in water, specifically includes the following steps:

[0011] S1, Prepare a solution to simulate natural water samples, select a mixed solution of glucose and sodium humate for COD calibration, and use it as the solution for detecting COD parameter of water quality by laser-induced fluorescence imaging method;

[0012] S2, Use laser-induced fluorescence technology to excite the water sample, obtain the fluorescence image by an image sensor, and perform color channel separation, feature extraction and analysis processing on the fluorescence image in combination with image processing technology;

[0013] S3, Analyze the RGB color characteristic values of the selected fluorescence region, adopt the partial least squares regression analysis method with single dependent variable, construct a COD concentration prediction model based on the average value of RGB color characteristics of the characteristic region of the fluorescence image, and realize the detection of COD water quality parameter.

[0014] In step S1, the ratio of sodium humate to glucose solution in the detection solution is determined by the COD value of the actual water sample for calibration.

[0015] Furthermore, use the system for detecting COD parameter of water quality by laser-induced fluorescence imaging method to detect and analyze water samples with a fixed range of COD concentration values, and set a fixed laser power.

[0016] In step S2, the fluorescence image is subjected to color channel separation, feature extraction, and analysis processing in combination with image processing technology, including: the computer device separates the RGB color channels of the collected fluorescence image ROI and calculates the regional average value of each channel. The calculation formula is as follows:

[0017]

[0018] In the formula, avg_R is the regional average value of the red channel, avg_G is the regional average value of the green channel, avg_B is the regional average value of the blue channel, y is the ordinate value of the pixel point, y st is the ordinate value of the starting pixel point of the ROI area, x is the abscissa value of the pixel point, x st is the abscissa value of the starting pixel point of the ROI area, H is the height of the ROI, W is the width of the ROI, R(x, y) is the pixel value of the red channel at this pixel position, G(x, y) is the pixel value of the green channel at this pixel position, and B(x, y) is the pixel value of the blue channel at this pixel position.

[0019] In step S3, a COD concentration prediction model based on the RGB color feature average value of the fluorescence image feature area is constructed, including: using partial least squares regression with a single dependent variable for multivariate statistical analysis. Therefore, the working regression equation of the COD concentration obtained based on the partial least squares method with a single dependent variable is:

[0020] C COD = k + a×R + b×G + c×B

[0021] In the formula, C COD is the original concentration value, a, b, and c are all regression coefficients, k is a constant term, R is the average value of the R value of the feature area, G is the average value of the G value of the feature area, and B is the average value of the B value of the feature area;

[0022] The working regression equation of the COD concentration obtained based on the partial least squares method with a single dependent variable is a PLSR regression model established based on the image feature values RGB.

[0023] Another object of the present invention is to provide a water quality COD parameter detection system by laser-induced fluorescence imaging method. This system is realized by the above-mentioned water quality COD parameter detection method by laser-induced fluorescence imaging. This system includes:

[0024] A laser emission module, which is provided with a laser, a first converging lens, and a power supply module. The laser uses a purple semiconductor laser. The laser beam emitted by the laser irradiates the detected water sample through the first converging lens to induce fluorescence, and the power supply module supplies power to the laser;

[0025] An image acquisition module that uses an image sensor to capture the fluorescence signal excited by the water sample and acquire a fluorescence image. A second converging lens is placed in front of the image sensor, and a filter is arranged behind the second converging lens.

[0026] A data processing module that uses a computer device to perform operations such as color channel separation, feature extraction, and analysis on the fluorescence image data, constructs a COD concentration regression model based on the RGB color feature values extracted from the fluorescence image, and calculates the COD prediction value of the water sample.

[0027] A sample module composed of a four-way optical quartz cuvette and a water sample. The output laser passes through the cuvette containing the water sample, and the image sensor acquires the fluorescence image in the vertical direction of the laser incidence.

[0028] An upper computer display module for real-time display of the detected value of the COD concentration.

[0029] Furthermore, the image sensor adopts a CCD image sensor or a CMOS image sensor; the computer device adopts one or more of a PC, a single-chip microcomputer, a Raspberry Pi, or an FPGA.

[0030] Furthermore, the computer device is responsible for the control of the laser, and adjusts the exposure time, color gain, and saturation parameters of the image sensor according to the intensity and saturation level of the fluorescence signal.

[0031] Furthermore, the computer device sets the pulse frequency and duty cycle through the laser emission module to achieve pulse control of the laser, receives the RGB image collected by the image sensor, and performs operations such as color channel separation, feature extraction, and analysis; and conducts data communication with the upper computer through wired or wireless means to achieve real-time display of the COD detection value.

[0032] Combining all the above technical solutions, the beneficial effects of the present invention are as follows:

[0033] First, the chemical oxygen demand (COD) is an important indicator for evaluating the degree of organic pollution in water bodies. The new method for detecting COD in water based on laser-induced fluorescence imaging proposed in the present invention uses an image sensor to detect the fluorescence image information generated by organic matter in water under ultraviolet laser excitation. After image processing, the COD value corresponding to the concentration of organic matter in water is obtained. In the experiment, a mixed solution of sodium humate and glucose with different concentrations was detected, and the fluorescence images corresponding to the detection solutions with different concentrations were obtained. The partial least squares regression analysis method was used to construct a prediction model for the COD concentration value based on the average value of the RGB color characteristics of the characteristic region of the fluorescence image. The experiment shows that it has the advantages of high sensitivity, strong stability, miniaturization, and non-contact detection when detecting low-concentration COD, and is suitable for rapid real-time water quality monitoring. With the advantages of high sensitivity, strong stability, miniaturization, and non-contact detection when detecting low-concentration COD, it is suitable for rapid real-time water quality monitoring and is expected to be widely used in the field of water quality monitoring.

[0034] Second, the present invention combines the advantages of fluorescence method and image processing, and proposes to use fluorescence imaging method to detect the water quality COD parameter. In the experiment, the present invention uses laser-induced fluorescence technology to excite the water sample, obtains the fluorescence image by an image sensor, and then combines image processing technology to extract and process the features of the fluorescence image. By analyzing the RGB color characteristic values of the selected fluorescence region, a COD concentration prediction model is constructed to achieve accurate, non-contact and rapid detection of the COD water quality parameter.

[0035] Third, the present invention constructs a laser-induced fluorescence imaging COD detection system. The system uses the partial least squares regression analysis method with a single dependent variable to construct a prediction model for the COD concentration value based on the average value of the RGB color characteristics of the characteristic region of the fluorescence image. It is suitable for detecting low-concentration organic matter in water, and has the advantages of high sensitivity, strong stability, non-contact and fast detection speed, and has good application value in rapid water quality monitoring and early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings here are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure;

[0037] Figure 1 is the flow chart of the method for detecting the water quality COD parameter by the laser-induced fluorescence imaging method provided by the embodiment of the present invention;

[0038] Figure 2 is the system diagram of the method for detecting the water quality COD parameter by the laser-induced fluorescence imaging method provided by the embodiment of the present invention;

[0039] Figure 3 is the structural diagram of the system hardware principle provided by the embodiment of the present invention;

[0040] In the figure: 1. Laser emission module; 1-1. Laser; 1-2. First converging lens; 1-3. Power supply module; 2. Image acquisition module; 2-1. Image sensor; 2-2. Second converging lens; 2-3. Filter; 3. Data processing module; 4. Sample module; 4-1. Colorimetric cuvette; 4-2. Water sample; 5. Host computer display module. Specific embodiments

[0041] To make the above objects, features, and advantages of the present invention more obvious and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description in order to fully understand the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0042] The main advantages of fluorescence analysis are high sensitivity and non-contact detection, while image processing technology not only can perform non-contact detection but also has advantages such as a large field of view and low cost. The innovation of the present invention lies in: the present invention combines the advantages of fluorescence method and image processing, and proposes a laser-induced fluorescence imaging method to achieve the detection of low-concentration COD parameters in water quality. By analyzing the color characteristic values and their changes of the fluorescence images of COD detection samples with different concentrations, a partial least squares regression analysis method with a single dependent variable is used to construct a prediction model for the COD concentration value based on the average RGB color characteristics of the characteristic regions of the fluorescence images. The laser-induced fluorescence imaging method COD detection system designed by the present invention is applicable to the detection of low-concentration organic substances in water, and has the advantages of high sensitivity, strong stability, non-contact, and fast detection speed, and has good application value in rapid water quality monitoring and early warning.

[0043] Example 1, as Figure 1 shown, the laser-induced fluorescence imaging method for water quality COD parameter detection provided by the embodiment of the present invention includes the following steps:

[0044] S1. Prepare a solution to simulate natural water samples, and select a mixed solution of glucose and sodium humate for COD calibration as the water quality COD parameter detection solution for the laser-induced fluorescence imaging method;

[0045] S2. Use laser-induced fluorescence technology to excite the water sample 4-2, obtain the fluorescence image by the image sensor 2-1, and perform color channel separation, feature extraction, and analysis processing on the fluorescence image in combination with image processing technology;

[0046] S3. Analyze the RGB color characteristic values of the selected fluorescence region, and adopt the partial least squares regression analysis method with a single dependent variable to construct a COD concentration prediction model based on the average RGB color characteristics of the characteristic region of the fluorescence image, so as to realize the detection of COD water quality parameters.

[0047] In step S1, a specific ratio of sodium humate and glucose solution in the detection solution is adopted. This specific ratio is obtained by the applicant through repeated and rigorous experimental optimization and screening. Through practical verification, when the solution mixed in this ratio is applied to the COD detection process, it can accurately and stably play its role.

[0048] In step S2, use the water quality COD parameter detection system based on laser-induced fluorescence imaging method to detect and analyze the water sample 4-2 with COD concentration value.

[0049] The STM32 microcontroller separates the RGB color channels of the collected fluorescence image ROI and calculates the regional average value of each channel. The calculation formula is as follows:

[0050]

[0051] In the formula, avg_R is the average value of the red channel region, avg_G is the average value of the green channel region, avg_B is the average value of the blue channel region, y is the ordinate value of the pixel point, y st is the ordinate value of the starting pixel point of the ROI region, x is the abscissa value of the pixel point, x st is the abscissa value of the starting pixel point of the ROI region, H is the height of the ROI, W is the width of the ROI, R(x, y) is the pixel value of the red channel at this pixel position, G(x, y) is the pixel value of the green channel at this pixel position, B(x, y) is the pixel value of the blue channel at this pixel position, the image storage format is RGB565, and the data value range is 0-255.

[0052] In step S3, the partial least squares regression with a single dependent variable is adopted for multivariate statistical analysis. Therefore, the working regression equation of COD concentration obtained according to the partial least squares method with a single dependent variable is:

[0053] C COD = k + a×R + b×G + c×B

[0054] In the formula, C COD is the original concentration value, a, b, and c are all regression coefficients, k is the constant term, R is the average value of the R value of the characteristic region, G is the average value of the G value of the characteristic region, B is the average value of the B value of the characteristic region;

[0055] In the present invention, the value of the constant term k is -2.1562, the regression coefficient a is 0.1013, the regression coefficient b is 0.0709, and the regression coefficient c is 0.0479. Then, we have:

[0056] C COD = -2.1562 + 0.1013×R + 0.0709×G + 0.0479×B

[0057] The working regression equation of COD concentration obtained by the partial least squares method with a single dependent variable is a PLSR regression model established based on the image feature values RGB. In the present invention, the coefficient of determination R 2 = 0.9975, and the root mean square error RMSE = 0.1865, indicating that the model has a very high degree of fitting to the data and has a high prediction accuracy. To more intuitively observe the reliability of the regression equation, the relative errors between the original concentration values of each working point and the model predictions are calculated. The results show that in the lower concentration range, the relative error of the prediction model remains below 10%.

[0058] Example 2, as Figure 3 shown, the laser-induced fluorescence imaging method for water quality COD parameter detection system provided by the embodiment of the present invention is composed of a laser emission module 1, an image acquisition module 2, a data processing module 3, a sample module 4, and a host computer display module 5.

[0059] As Figure 2 shown, the laser emission module 1 includes a laser 1-1, a first converging lens 1-2, and a laser power supply; among them, the laser 1-1 is a 405nm purple semiconductor laser 1-1, which vertically irradiates the detected water sample 4-2 to induce fluorescence. The laser beam emitted by the laser 1-1 passes through the first converging lens 1-2 to irradiate the detected water sample 4-2 to induce fluorescence, and the power supply module 1-3 supplies power to the laser 1-1;

[0060] The image acquisition module 2 uses an image sensor 2-1 to capture the fluorescence signal excited by the water sample 4-2 and collect the fluorescence image; a second converging lens 2-2 is provided at the front end of the image sensor 2-1, and a long-wave filter 2-3 is provided behind the second converging lens 2-2. The image acquisition module 2 uses an image sensor 2-1 to capture the fluorescence signal excited by the water sample 4-2 and collect the fluorescence image. To ensure clear imaging, a second converging lens 2-2 with a focal length of 12mm is placed at the front end of the image sensor 2-1; a long-wave filter 2-3 is placed behind the second converging lens 2-2 to reduce the influence of laser scattering on the fluorescence image. The image sensor 2-1 can use a CCD image sensor and a CMOS image sensor. The computer device can select a 32-bit single-chip microcomputer, a 51-bit single-chip microcomputer, a Raspberry Pi, an FPGA, etc., and the selection is determined according to the specific application scenario and technical requirements.

[0061] In the data processing module 3, the computer device serves as the core data processing unit, collecting RGB image data, performing color channel separation, feature extraction, and analysis operations on the fluorescence image data, constructing a COD concentration regression model based on the RGB color feature values extracted from the fluorescence image, and finally calculating the COD prediction value of the water sample 4-2. In addition, the microcontroller is responsible for controlling the laser 1-1 and can also adjust parameters such as the exposure time, color gain, and saturation of the image sensor 2-1 according to the intensity and saturation level of the fluorescence signal.

[0062] The sample module 4 consists of a four-way optical quartz cuvette 4-1 and a water sample 4-2. The output laser passes through the cuvette 4-1 containing the water sample 4-2, and the image sensor 2-1 collects the fluorescence image in the vertical direction of the laser incidence.

[0063] The upper computer display module 5 is used to display the detected value of the COD concentration in real time.

[0064] The laser-induced fluorescence imaging method for water quality COD parameter detection system provided by the embodiment of the present invention selects the STM32 microcontroller as the MCU. The STM32 realizes the pulse control of the laser 1-1 through the laser emission module 1 and receives the RGB image collected by the image sensor 2-1. Data communication is carried out between the STM32 and the upper computer through the Bluetooth module to realize the real-time display of the COD detection value.

[0065] The system also includes a storage module and a power supply terminal. The storage module is mainly used to store the fluorescence image and the extracted regional RGB color feature values to ensure the real-time and rapid refresh of the image and data. The power supply terminal is responsible for ensuring the normal operation of the entire system. This system combines laser-induced fluorescence and image processing to detect organic matter in water, and uses the fluorescence phenomenon generated by the laser vertically irradiating the detection solution to realize the detection of COD concentration.

[0066] By preparing a mixed solution of glucose solution and sodium humate solution as the COD detection solution for the laser-induced fluorescence imaging method, the specific ratio of sodium humate to glucose solution in the detection solution is adopted. This specific ratio is obtained by the applicant through repeated and a large number of rigorous experimental optimization and screening. Through practical verification, when the solution mixed in this ratio is applied to the COD detection process, it can accurately and stably play its effectiveness.

[0067] The present invention uses partial least squares regression (PLSR) with a single dependent variable for multivariate statistical analysis. PLSR can effectively extract information from multiple related independent variables and construct an accurate prediction model for the COD concentration value. In addition, PLSR also has the function of data dimensionality reduction, which can eliminate the noise interference in the data, thereby improving the stability of the prediction model.

[0068] There is a strong correlation between the fluorescence signal intensity of COD detection samples with different concentrations and the changes in the RGB color characteristic values of their fluorescence images, verifying the feasibility and accuracy of the rapid water quality COD detection method based on the laser-induced fluorescence imaging method proposed in the present invention under low concentration conditions. The present invention proposes a new method for detecting COD in water based on laser-induced fluorescence imaging. The CMOS detection module in the image sensor 2-1 is used to detect the fluorescence image information of organic substances. Combining image processing and regression analysis methods, an innovative detection system is constructed. There has been no application of a similar mature and comprehensively advantageous technical solution before, opening up a new path for COD detection. The present invention innovatively introduces laser-induced fluorescence imaging, fully excavates image information, and proves that even in the face of water bodies with complex compositions, by using fluorescence image features combined with intelligent algorithm processing, the COD value can be accurately obtained.

[0069] As described above, it is only a relatively preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any modification, equivalent replacement, and improvement made by those skilled in the art within the technical scope disclosed by the present invention and within the spirit and principle of the present invention should be covered by the protection scope of the present invention.

Claims

1. A method for detecting COD parameters of water quality by laser induced fluorescence imaging, characterized in that: The method combines fluorescence analysis and image processing technology to detect COD concentration in water, and specifically includes the following steps: S1, preparing a solution to simulate a natural water sample, selecting a mixed solution of glucose and sodium humate for COD calibration, and using it as a water quality COD parameter detection solution for laser induced fluorescence imaging; S2, using laser induced fluorescence technology to excite the water sample (4-2), using the image sensor (2-1) to obtain a fluorescence image, and combining image processing technology to perform color channel separation, feature extraction and analysis processing on the fluorescence image; S3, analyze the RGB color feature values ​​of the selected fluorescence area, use the partial least squares regression analysis method of a single dependent variable, and build a COD concentration prediction model based on the average value of the RGB color features of the characteristic area of ​​the fluorescence image to realize the detection of COD water quality parameters.

2. The method for detecting COD parameters of water quality by laser induced fluorescence imaging according to claim 1, characterized in that: In step S1, the ratio of sodium humate to glucose solution in the test solution is determined by calibrating the COD value of the actual water sample (4-2).

3. The method for detecting COD parameters of water quality by laser induced fluorescence imaging according to claim 1, characterized in that: The water quality COD parameter detection system using laser induced fluorescence imaging was used to detect and analyze water samples (4-2) with a fixed range of COD concentration values, and a fixed laser power was set.

4. The method for detecting COD parameters of water quality by laser induced fluorescence imaging according to claim 1, characterized in that: In step S2, the color channel separation, feature extraction and analysis processing of the fluorescence image are performed in combination with image processing technology, including: the computer device performs RGB color channel separation on the ROI of the collected fluorescence image, and calculates the regional average value of each channel. The calculation formula is as follows: Where avg_R is the average value of the red channel area, avg_G is the average value of the green channel area, avg_B is the average value of the blue channel area, y is the vertical coordinate value of the pixel point, and y st is the vertical coordinate value of the starting pixel point of the ROI area, x is the horizontal coordinate value of the pixel point, and x st is the horizontal coordinate value of the starting pixel of the ROI area, H is the ROI height, W is the ROI width, R(x,y) is the pixel value of the red channel at the pixel position, G(x,y) is the pixel value of the green channel at the pixel position, and B(x,y) is the pixel value of the blue channel at the pixel position.

5. The method for detecting COD parameters of water quality by laser induced fluorescence imaging according to claim 1, characterized in that: In step S3, a COD concentration prediction model based on the average value of the RGB color characteristics of the characteristic area of ​​the fluorescent image is constructed, including: using partial least squares regression of a single dependent variable for multivariate statistical analysis, so the COD concentration working regression equation obtained by partial least squares method of a single dependent variable is: C COD =k+a×R+b×G+c×B In the formula, C COD is the original concentration value, a, b, c are regression coefficients, k is a constant term, R is the average R value of the characteristic area, G is the average G value of the characteristic area, and B is the average B value of the characteristic area; The working regression equation of COD concentration obtained by partial least squares method of single dependent variable is a PLSR regression model established based on image eigenvalue RGB.

6. A water quality COD parameter detection system using laser induced fluorescence imaging, characterized in that: The system is implemented by the water quality COD parameter detection method using the laser induced fluorescence imaging method according to any one of claims 1 to 5, and the system comprises: A laser emission module (1) is provided with a laser (1-1), a first converging lens (1-2) and a power module (1-3); the laser (1-1) adopts a violet semiconductor laser; a laser beam emitted by the laser (1-1) irradiates a detection water sample (4-2) through the first converging lens (1-2) to induce fluorescence; and the power module (1-3) supplies power to the laser (1-1); An image acquisition module (2) uses an image sensor (2-1) to capture a fluorescent signal stimulated by a water sample (4-2) and collect a fluorescent image; a second converging lens (2-2) is placed at the front end of the image sensor (2-1), and a filter (2-3) is arranged behind the second converging lens (2-2); The data processing module (3) uses a computer device to perform color channel separation, feature extraction and analysis processing operations on the fluorescent image data, constructs a COD concentration regression model based on the RGB color feature values ​​extracted from the fluorescent image, and calculates the COD prediction value of the water sample (4-2); The sample module (4) is composed of a four-way quartz cuvette (4-1) and a water sample (4-2); the output laser passes through the cuvette (4-1) containing the water sample (4-2), and the image sensor (2-1) collects a fluorescent image in a direction perpendicular to the incident laser; The upper computer display module (5) is used to display the detected value of COD concentration in real time.

7. The laser induced fluorescence imaging water quality COD parameter detection system according to claim 6, characterized in that: The image sensor (2-1) adopts a CCD image sensor or a CMOS image sensor; the computer device adopts one or more of a PC, a single-chip microcomputer, a Raspberry Pi or an FPGA.

8. The laser induced fluorescence imaging water quality COD parameter detection system according to claim 7 is characterized in that: The computer device is used to control the laser (1-1) and adjust the exposure time, color gain and saturation parameters of the image sensor (2-1) according to the intensity and saturation level of the fluorescent signal.

9. The laser induced fluorescence imaging water quality COD parameter detection system according to claim 8, characterized in that: The computer device sets the pulse frequency and duty cycle through the laser emission module (1) to realize pulse control of the laser (1-1), receives the RGB image collected by the image sensor (2-1), and performs color channel separation, feature extraction and analysis and processing operations; and communicates data with the host computer through wired or wireless means to realize real-time display of COD detection values.

Images