Portable seawater chlorophyll remote sensing monitoring device and method

Abstract

The invention belongs to the technical field of spectral component measurement, and particularly discloses a portable seawater chlorophyll remote sensing monitoring device and method.A 365 nm LED light source and a 405 nm LED light source are adopted, and a collimating mirror, an adjustable attenuation piece and a reflecting mirror are arranged behind each light source; light emitted by the two light sources is scanned by the galvanometer and is emitted into water; chlorophyll radiation fluorescence in water enters the slit, is dispersed to the transmission grating by the prism after passing through the slit, and is converged to the CMOS camera by the imaging lens to obtain a fluorescence spectrum. The light sources of 365 nm and 405 nm are adopted to be well matched with the absorption spectrum of chlorophyll, and fluorescence of chlorophyll can be efficiently excited. By establishing a relation model of chlorophyll concentration and a radiated fluorescence spectrum, the chlorophyll concentration is inversed by using a measured fluorescence spectrum signal in practice. The chlorophyll in the water is excited by utilizing light with two wavelengths to obtain multiple groups of fusion fluorescence spectrums, and the chlorophyll content in the water is accurately quantified by combining a relation model.

Description

Technical Field

[0001] The present invention relates to a non-contact monitoring device for seawater chlorophyll, and also relates to a method for monitoring seawater chlorophyll concentration. Background Art

[0002] Enteromorpha prolifera green tide is one of the common marine ecological disasters. Its outbreak will not only cause serious impacts on the marine ecological environment, but also bring huge losses to fishery resources, tourism, etc. Monitoring the content of chlorophyll in water can accurately and timely warn of the Enteromorpha prolifera green tide, which is of great significance for preventing and controlling the Enteromorpha prolifera green tide disaster, protecting the marine ecological environment, and maintaining the stable development of the marine economy. The current in-water chlorophyll monitors need to be placed in water and continuously monitor the long-term state of the area. For large-area water monitoring, a large number of devices need to be managed and there is a certain amount of loss. Therefore, there is a greater need for some non-contact remotely detectable devices that can be measured as the mobile vehicle moves. This not only has a large measurement patrol area and high real-time performance, but also can greatly reduce costs.

[0003] CN114646621B discloses a leaf in-situ dark adaptation device and method for chlorophyll fluorescence monitoring. This patent needs to use sunlight to induce chlorophyll to produce fluorescence, relying on external light, and the signal is relatively weak compared to laser-induced fluorescence and LED-induced fluorescence signals.

[0004] CN116067930A discloses a non-contact chlorophyll monitoring device, which adopts a detection method combining an ultraviolet light excitation light source and a detector group. Due to the use of a filter, the detection range of the chlorophyll fluorescence spectrum of this device is limited. Traditional photo-induced fluorescence devices usually have only one excitation light source, and the processing after obtaining the chlorophyll fluorescence spectrum shape is relatively simple. Therefore, these devices have certain limitations in detecting the composition and content of chlorophyll. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention provides a device and method for monitoring the content of chlorophyll in water with a dual-wavelength light source. The device includes a dual-wavelength emission system, a galvanometer scanner, a slit, a prism, a transmission grating, an imaging lens, and a CMOS camera. The dual-wavelength emission system includes a 365nm LED light source, a 405nm LED light source, a collimating mirror, an adjustable attenuator, a reflecting mirror, etc. The method uses this device to obtain the chlorophyll fluorescence spectrum through dual LEDs, and compares the spectrum shape and intensity with the prepared standard chlorophyll solution to achieve the detection of chlorophyll concentration.

[0006] The technical solution of the present invention is: a portable remote sensing monitoring device for seawater chlorophyll, the device includes a dual-wavelength emission system, a galvanometer scanner, a slit, a prism, a transmission grating, an imaging lens, and a CMOS camera; The dual-wavelength emission system includes LED light sources of two different wavelengths. On the optical path of each LED light source, a collimating mirror, an adjustable attenuator, and a reflecting mirror are sequentially arranged; The light emitted by the two LED light sources passes through galvanometer scanning after passing through the emission system and then enters the water; The fluorescence radiated by chlorophyll in the water enters the slit of the device, is dispersed by a prism to a transmission grating after passing through the slit, and then is converged to a CMOS camera by an imaging lens.

[0007] Further, one of the two LED light sources of different wavelengths is a 365 nm light source, and the other is a 405 nm light source.

[0008] Further, the light emitted by the two LED light sources enters the galvanometer along the same path after passing through the reflecting mirror.

[0009] Further, a first through hole and a second through hole are provided on the device. The light after galvanometer scanning enters the water through the first through hole, and the fluorescence radiated by chlorophyll in the water enters the slit through the second through hole.

[0010] The present invention provides a method for realizing the monitoring of chlorophyll in water, based on the device: First, establish a relationship model between the chlorophyll concentration in water and the fluorescence spectrum of radiation, and then invert the concentration of chlorophyll through the fluorescence spectrum signal measured actually; Among them, the method for establishing the relationship model between the chlorophyll concentration in water and the fluorescence spectrum of radiation is: 1) First, give a series of reference values of chlorophyll concentration in water; 2) At each reference concentration, obtain two sets of original fluorescence spectrum signal intensities through the dual-wavelength emission system; 3) Combine the two sets of signals into a spectral matrix, that is, the fusion spectrum of the two excitation light sources : ;

[0011] In the formula, is the fluorescence spectrum signal intensity of wavelength excited by the first light source, is the fluorescence spectrum signal intensity of wavelength excited by the second light source, and so on; 4) Adjust the adjustable attenuators of the two light sources p times to obtain the fusion spectra of p times as: ;

[0012] 5) The spectral intensity after SG smoothing filtering and normalization is: ;

[0013] 6) Perform two-dimensional convolution calculation on to obtain: ;

[0014] where the convolution kernel has a size of , H and W are the height and width of the convolution kernel, extracting local features, i and j respectively represent the i th row and the j th column, and the bias term B is used to adjust the activated output; 7) Use ReLU activation to obtain matrix : ;

[0015] 8) Perform max pooling to obtain the downsampled matrix : ;

[0016] 9) Flatten the data after max pooling to obtain a one-dimensional vector : ;

[0017] 10) Calculate the regression concentration through the fully connected layer: ;

[0018] In the formula, represents the concentration, is the weight of the output layer, is the bias of the output layer; Using the above relationship, the mapping relationship between any concentration and fluorescence spectrum is trained. In practice, the measured spectral data is input into the model of the mapping relationship to inversely calculate the concentration of chlorophyll.

[0019] The present invention adopts the above technical solutions, and compared with the prior art, it shows the following remarkable technical effects: 1. By setting up a monitoring device including a dual-wavelength emission system, the present invention excites the water body to generate chlorophyll fluorescence through 365nm LED and 405nm LED light sources, and adjusts the luminous intensity of the two-wavelength LEDs through an adjustable attenuation sheet to obtain different fusion spectra, and can accurately obtain chlorophyll concentration information, and the device has a low cost.

[0020] 2. In the present invention, chlorophyll in water is excited by light of two wavelengths, and the generated fluorescence spectra have certain differences. This difference can be utilized. Meanwhile, by obtaining multiple sets of fused fluorescence spectra under the excitation intensity and combining with an artificial intelligence algorithm, the chlorophyll concentration in water can be accurately quantified. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a structural diagram of the monitoring device of the present invention; Figure 2 They are fluorescence spectra obtained by chlorophyll of different concentrations under two excitation wavelengths. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] The present invention will be described in detail below in conjunction with embodiments. Those skilled in the art should know that the following embodiments are not the only limitations on the technical solution of the present invention. Any equivalent transformation or modification made under the spirit of the technical solution of the present invention should be regarded as belonging to the protection scope of the present invention.

[0023] As Figure 1 shown, the present invention provides a monitoring device for chlorophyll content in water with two wavelengths. The device is portable and can be measured along with mobile tools in water. The device includes a two-wavelength emission system, a galvanometer, a housing, a through hole, a slit, a prism, a transmission grating, an imaging lens, and a CMOS camera.

[0024] The two-wavelength emission system includes an LED power supply and two LED light sources. One of the two LED light sources uses a 365nm LED light source, and the other uses a 405nm LED light source. The two light sources are powered by the same LED power supply; a collimating mirror, an adjustable attenuator, and a reflecting mirror are arranged on the optical path of each LED light source. As Figure 1 shown, a collimating mirror 1, an adjustable attenuator 1, and a reflecting mirror 1 are sequentially arranged on the optical path of the 365nm LED light source, and a collimating mirror 2, an adjustable attenuator 2, and a reflecting mirror 2 are sequentially arranged on the optical path of the 405nm LED light source. The light emitted by the reflecting mirror 1 and the reflecting mirror 2 is reflected again by the reflecting mirror 3 and reflected onto the galvanometer.

[0025] The LED power supply can sequentially control the 365nm LED light source and the 405nm LED light source to emit horizontal excitation light. The 365nm LED light source is collimated by the collimating mirror 1 to the adjustable attenuator 1, then reflected to the vertical direction by the reflecting mirror 1, transmitted through the reflecting mirror 2 to the reflecting mirror 3, and then totally reflected onto the galvanometer; the 405nm LED light source is collimated by the collimating mirror 2 to the adjustable attenuator 2, then to the reflecting mirror 2 and reflected by it to the reflecting mirror 3, and then totally reflected onto the galvanometer.

[0026] The adjustable attenuation sheet can make the intensities of the two LED excitation lights adjustable. When the light intensity ratio of 365 nm to 405 nm is 1:1, 1:2, or 2:1, multiple sets of different fluorescence emission spectra can be obtained.

[0027] The first mirror is a single-sided total reflection mirror that totally reflects the light emitted by the 365 nm LED light source. The second mirror has a total reflection film on one side and an anti-reflection film on the other side. The side with the total reflection film is used to totally reflect the light emitted by the 405 nm LED light source, and the side with the anti-reflection film is used to transmit the light from the first mirror. The third mirror is also a single-sided total reflection mirror that is used to totally reflect the light from the first and second mirrors. The third mirror can change the direction of the light so that it hits the galvanometer mirror. If the first and second mirrors can directly irradiate the light onto the galvanometer mirror through reasonable layout, it is also possible not to use this third mirror.

[0028] The mirrors are set up so that the light emitted by the 365 nm LED light source and the 405 nm LED light source can finally reach the galvanometer mirror along the same optical path, and then be reflected by the galvanometer mirror through the first through-hole. The light from the two optical path systems finally enters the galvanometer mirror. The galvanometer mirror can be controlled by a galvanometer control system for one-dimensional scanning. The scanned light can be emitted through the first through-hole into the water body containing chlorophyll. The fluorescence radiated by chlorophyll in the water enters the slit through the second through-hole. The first through-hole and the second through-hole are constructed on the housing of the device, and all other components are arranged inside the housing.

[0029] The fluorescence radiated by chlorophyll in the water enters the slit through the second through-hole. The slit can easily separate the spectrum. The light after passing through the slit is dispersed by the prism to the transmission grating. The prism and the grating play a role in dispersion. The prism and the grating can be integrally formed. The grating is formed by engraving on the inclined surface of a right-angled glass prism, forming a combined configuration of the prism and the grating. The diffracted light of the transmission grating is converged by the imaging lens to the photosensitive surface of the CMOS camera. The image obtained by the CMOS camera can be collected by a computer and processed into a two-dimensional spectrogram.

[0030] The present invention uses an LED light source. The LED light source has a lower cost, lower power consumption compared to a laser or a traditional light source, can work continuously for a long time, and is small in size and suitable for portable monitoring.

[0031] The present invention selects light sources with wavelengths of 365 nm and 405 nm because the excitation in the blue-violet light band matches well with the absorption spectrum of chlorophyll a, which helps to efficiently excite the fluorescence of chlorophyll.

[0032] The present invention uses a galvanometer mirror, which can scan the light beam in the horizontal direction and is more helpful for rapid large-area monitoring compared to single-point detection.

[0033] Using this set of devices of the present invention, two different chlorophyll radiation fluorescence spectra can be obtained under the excitation of two light sources with different wavelengths. Since it is double-wavelength light excitation, the fluorescence spectra generated by the excitation of the two wavelengths are somewhat different, and this difference can be used to accurately quantify the chlorophyll concentration in water. Usually, when measuring the concentration of a certain water area, the two light sources are excited simultaneously, or they can also be excited successively. Figure 2 The fluorescence spectra obtained for the same concentration of chlorophyll under two excitation wavelengths are somewhat different.

[0034] The present invention provides a method for monitoring the chlorophyll content in water. The core idea is to first establish a relationship model between the concentration of chlorophyll in water and the radiation fluorescence spectrum, and then invert the concentration of chlorophyll through the captured fluorescence spectrum signal in actual measurement. Therefore, the key of the present invention lies in providing the establishment of the inversion model.

[0035] The fluorescence spectrum radiated by chlorophyll in water usually ranges from 620 nm to 800 nm. The fluorescence emission peak of chlorophyll a is usually around 680 nm, and there is also a secondary peak at 730 nm. By setting a parameter in the range of 0 to 1 to control the adjustable attenuation filter to control the intensity of the 365 nm light and 405 nm light irradiated onto the seawater, a set of fused spectrum data at a certain concentration will be obtained. Prepare chlorophyll with multiple concentrations, obtain the fused spectra at these concentrations, and process these data and store them for future use.

[0036] Let and be a set of original fluorescence spectrum signal intensities obtained when two light sources, LED1 and LED2, excite different lights (in the present invention, LED1 and LED2 are 365 nm and 405 nm light sources respectively): ;

[0037] ;

[0038] In the formula, is the spectral signal intensity of the wavelength obtained by the excitation of the LED1 light source, and so on for the others.

[0039] Combine the two signals to form a new spectral matrix, that is, the fused spectrum of the two excitation light sources : ;

[0040] By adjusting the adjustable attenuation filters of the two light sources p times, p times of fused spectra can be obtained as: ;

[0041] The spectral intensity after SG smoothing filtering and normalization is: ;

[0042] To effectively extract features from the two-dimensional spectral intensity matrix, perform two-dimensional convolution calculation on to obtain: ;

[0043] where the convolution kernel has a size of , H and W are the height and width of the convolution kernel, respectively, to extract local features. i and j represent the i th row and the j th column, respectively, and the bias term B is used to adjust the activated output.

[0044] To enhance the network's expressive ability, introduce non-linear activation. Use the most common ReLU activation to obtain the matrix : ;

[0045] Then perform downsampling to reduce parameters, which can retain key features and improve generalization. Perform max pooling to obtain the downsampled matrix : ;

[0046] To input these extracted spatial features as overall information into the regression model, establish a mapping model for all features, and flatten the data after max pooling to obtain a one-dimensional vector : ;

[0047] Then establish a non-linear mapping between the input features and the output target, and calculate the regression concentration through the fully connected layer: ;

[0048] In the formula, represents the concentration, is the weight of the output layer, and is the bias of the output layer.

[0049] The above process only obtains the spectrum of a certain concentration. Spectra at a series of concentrations can be obtained in the same way, and the mapping relationship between any concentration and spectrum can be obtained through a large number of spectral trainings.

[0050] Use the above method to train and store the spectra of different concentrations obtained in the laboratory. Input the newly measured spectral data into the trained model to inversely calculate the concentration of chlorophyll.

Claims

1. A portable remote sensing monitoring device for seawater chlorophyll, characterized in that: The device includes a dual-wavelength emission system, a galvanometer scanner, a slit, a prism, a transmission grating, an imaging lens, and a CMOS camera; The dual-wavelength emission system includes two LED light sources with different wavelengths. On the optical path of each LED light source, a collimating mirror, an adjustable attenuator, and a reflecting mirror are sequentially arranged; The light emitted by the two LED light sources passes through the emission system and is scanned by the galvanometer scanner, and then enters the water; The chlorophyll fluorescence in the water enters the slit of the device, is dispersed by the prism to the transmission grating after passing through the slit, and then is converged to the CMOS camera by the imaging lens.

2. The portable remote sensing monitoring device for seawater chlorophyll according to claim 1, characterized in that: The two LED light sources with different wavelengths are one 365nm light source and the other 405nm light source.

3. The portable seawater chlorophyll remote sensing monitoring device according to claim 1, wherein: The light emitted by the two LED light sources enters the galvanometer scanner along the same path after passing through the reflecting mirror.

4. The portable seawater chlorophyll remote sensing monitoring device according to claim 1, wherein: A through hole 1 and a through hole 2 are provided on the device; The light after the galvanometer scanner scans enters the water through the through hole 1; The fluorescence radiated by the chlorophyll in the water enters the slit through the through hole 2.

5. A method for monitoring chlorophyll in water implemented by the device according to any one of claims 1-4, characterized in that: First, establish a relationship model between the chlorophyll concentration in water and the fluorescence spectrum of radiation, and then invert the chlorophyll concentration through the actually measured fluorescence spectrum signal; Among them, the method for establishing the relationship model between the chlorophyll concentration in water and the fluorescence spectrum of radiation is: 1) First, give a series of reference values of chlorophyll concentration in water; 2) At each reference concentration, obtain two sets of original fluorescence spectrum signal intensities through the dual-wavelength emission system; 3) Combine the two sets of signals to form a spectral matrix, i.e., the fusion spectrum of the two excitation light sources : ； In the formula, is the fluorescence spectral signal intensity at the wavelength obtained by excitation of the first light source, is the fluorescence spectral signal intensity at the wavelength obtained by excitation of the second light source, and so on; 4) Adjust the adjustable attenuators of the two light sources p times, and the p times of fused spectra are: ； 5) The spectral intensity after SG smoothing filtering and normalization is: ； 6) Perform two-dimensional convolution calculation on to obtain: ； Among them, the convolution kernel has a size of , H and W being the height and width of the convolution kernel, extracts local features, i and j respectively represent the i th row and the j th column, and the bias term B is used to adjust the activated output; 7) Obtain a matrix using ReLU activation : ； 8) Perform max pooling to obtain the matrix after dimensionality reduction : ； 9) Flatten the data after max pooling to obtain a one-dimensional vector : ； 10) Calculate the regression concentration through the fully connected layer: ； wherein represents the concentration, is the weight of the output layer, is the bias of the output layer; Using the above relationship, train to obtain the mapping relationship between any concentration and fluorescence spectrum.

Images