Plankton in-situ imaging system and method applied in open scenes

Abstract

The present invention discloses a plankton in-situ imaging system and method applied in an open scene, the system comprises: a telecentric light source, a dichroic beam splitter, a back imaging module, a fluorescence imaging module, a control module and an image processing module; the telecentric light source is incident on the plankton plane in the water body, and the plankton target containing chlorophyll emits a longer wavelength fluorescence component under the excitation of the light emitted by the telecentric light source, which is mixed with the illumination light, and the dichroic beam splitter separates the two; the back imaging module performs back imaging on the separated illumination light, and the fluorescence imaging module performs fluorescence imaging on the separated fluorescence; the control module controls the back imaging module and the fluorescence imaging module to be triggered simultaneously, so as to realize the same frequency sampling of the backlight image and the fluorescence image of the same field of view. The present invention can realize the effective observation of micro-plankton.

Description

Technical Field

[0001] The present invention relates to the technical field of plankton monitoring, and in particular to an in-situ imaging system and method for plankton applied in an open scene. Background Art

[0002] As an important part of the marine ecosystem, accurate monitoring of plankton can provide an assessment basis for the health of the marine ecosystem. Imaging-based plankton observation methods can provide high-resolution morphological features of the target in a non-invasive manner. With the development of imaging devices and deep learning, high-frequency acquisition of high-resolution images and automatic classification and identification lay the foundation for real-time monitoring of plankton. However, existing open imaging platforms cannot achieve effective observation. Summary of the invention

[0003] The purpose of the present invention is to solve the technical problem of effectively observing micro-plankton and to propose an in-situ imaging system and method for plankton applied in an open scene.

[0004] The technical problem of the present invention is solved by the following technical solutions:

[0005] On the one hand, the present invention discloses an in-situ imaging system for plankton used in an open scene, comprising: a telecentric light source, a dichroic beam splitter, a back imaging module, a fluorescence imaging module, a control module and an image processing module; the telecentric light source is incident on a plankton plane in a water body, and a plankton target containing chlorophyll emits a longer wavelength fluorescence component under the excitation of light emitted by the telecentric light source, which is mixed with the illumination light, and the dichroic beam splitter separates the two; the back imaging module performs back imaging on the separated illumination light, and the fluorescence imaging module performs fluorescence imaging on the separated fluorescence; the control module controls the back imaging module and the fluorescence imaging module to be triggered simultaneously, so as to achieve the same frequency sampling of the backlight image and the fluorescence image of the same field of view.

[0006] In some embodiments, the image processing module is further used to perform motion blur identification and processing on the collected fluorescence image, and then fuse it with the background image to obtain a fused image.

[0007] In some embodiments, the back imaging module and the fluorescent imaging module include telecentric lenses, which cooperate with telecentric illumination to form a dual telecentric system.

[0008] In some embodiments, the back image imaging module includes a first telecentric lens and a CMOS camera. The back image imaging module uses the first telecentric lens to receive the illumination light reflected by the dichroic beam splitter and performs back image imaging on the CMOS camera.

[0009] In some embodiments, the fluorescence imaging module includes a filter, a second telecentric lens and a sCMOS camera. The fluorescence imaging module uses a filter to filter out wavelength components other than fluorescence, and performs fluorescence imaging on the sCMOS camera through the second telecentric lens.

[0010] In some embodiments, a scene classification data set is established through an image processing module, and a scene classification model is obtained by training based on a lightweight network using the scene classification data set. It is determined in an end-to-end manner whether a fluorescence image has motion blur. For fluorescence images with motion blur, the motion blur kernel is solved to perform deconvolution to eliminate the motion blur. The fluorescence image and the background image are fused at the pixel level in the spatial domain using a weighted averaging method to obtain a fused image that finally contains chlorophyll fluorescence information and sharp edges.

[0011] In some embodiments, by establishing a fluorescence illumination model, the illumination of the telecentric light source, the working distance of the telecentric light source and the telecentric lens are determined under the premise that the illumination of the image plane at which the fluorescence imaging module receives the fluorescence signal can be guaranteed under certain chlorophyll concentration conditions.

[0012] In some embodiments, the process of establishing the fluorescence illumination model includes: the process of the excitation light propagating to the target object, the fluorescence process of chlorophyll, and the process of the fluorescence emission light being received by the fluorescence imaging module.

[0013] In some embodiments, a sealed cabin is further included, wherein the back imaging module and the fluorescence imaging module are located in the sealed cabin; and the dichroic beam splitter is placed at the entrance of the sealed cabin, forming a 45-degree angle with the incident illumination light and fluorescence.

[0014] In some embodiments, the telecentric light source is a 460nm telecentric light source, which is incident on the plankton plane in the water body with uniform illumination through the sealed cabin.

[0015] Another aspect of the present invention discloses a method for in-situ imaging of plankton in an open scene, comprising the following steps:

[0016] S1. A telecentric light source is passed through a sealed cabin to illuminate the phytoplankton plane in the water body with uniform illumination. The chlorophyll-containing phytoplankton targets emit longer wavelength fluorescence components under the light excitation of the telecentric light source; the illumination light is transmitted and scattered by the phytoplankton and is mixed with the excited fluorescence;

[0017] S2, separating the overlapping illumination light and fluorescence;

[0018] S3, performing back imaging on the separated illumination light, and performing fluorescence imaging on the separated fluorescence;

[0019] S4. Perform the same frequency sampling on the backlight image and the fluorescence image of the same field of view.

[0020] In some embodiments, the step is further included: S5, performing motion blur identification and processing on the fluorescent image, and then fusing it with the background image to obtain a fused image.

[0021] In some embodiments, step S5 includes:

[0022] S51, establishing a scene classification data set, including: clear images, out-of-focus blurred images, and motion blurred images;

[0023] S52, training a scene classification model using the scene classification data set based on a lightweight network, and determining whether there is motion blur in the fluorescence image in an end-to-end manner;

[0024] S53, if there is no motion blur in the fluorescent image, directly proceed to step S44; if there is motion blur in the fluorescent image, solve the motion blur kernel and perform deconvolution to eliminate the motion blur;

[0025] S54, using a weighted average method to perform pixel-level fusion of the fluorescence image and the background image in the spatial domain, to obtain a fused image that finally contains chlorophyll fluorescence information and sharp edges.

[0026] In some embodiments, step S54 specifically includes: performing weighted averaging on each pixel of the fluorescent image M and the background image N, and the pixel value of the fused image G is obtained by performing weighted averaging on the pixel values ​​of M and N at corresponding positions, that is:

[0027] G(i,j)=w1M(i,j)+w2N(i,j)

[0028] Where G(i,j) is the pixel value at position (i,j) of the fused image, M(i,j) is the pixel value at position (i,j) of the fluorescence image, and N(i,j) is the pixel value at position (i,j) of the background image; w1 and w2 are weighting coefficients, and satisfy w1+w2=1;

[0029] The original background grayscale image is converted into an RGB three-channel color image, and in the R channel, the pixel values ​​of the fluorescence image are weightedly fused with the background image, so that the part containing chlorophyll appears red.

[0030] The beneficial effects of the present invention compared with the prior art include:

[0031] The plankton in-situ imaging system proposed in the present invention for use in open scenes is based on the microscopic back imaging optical path, and simultaneously performs chlorophyll fluorescence imaging on the target in the field of view, thereby achieving simultaneous back imaging and fluorescence imaging. The back imaging method can provide clear morphological features and high-contrast edges; chlorophyll fluorescence imaging can supplement the classification information for targets with insufficient morphological information, and distinguish between biological and non-biological components in particulate matter.

[0032] In some embodiments, the fluorescence image is fused with the background image after motion blur discrimination and processing. The resulting fused image increases the information dimension and can supplement chlorophyll information on the basis of morphological information, thereby improving target detection performance and the extraction of biological components of background particles. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 Schematic diagram of the structure of the plankton in-situ imaging system according to an embodiment of the present invention.

[0034] Figure 2 It is a working flow chart of the plankton in situ imaging system according to an embodiment of the present invention.

[0035] Figure 3a It is a fluorescence excitation spectrum diagram of Chlorella in the embodiment of the present invention.

[0036] Figure 3b It is a fluorescence excitation spectrum diagram of Chlorella in the embodiment of the present invention.

[0037] Figure 4 Schematic diagram of an image processing module according to an embodiment of the present invention.

[0038] Figure 5 This is a fluorescence image with motion blur collected in an embodiment of the present invention.

[0039] Figure 6 It is the back image collected in the embodiment of the present invention.

[0040] Figure 7 This is a fluorescent image after motion blur is eliminated in an embodiment of the present invention.

[0041] Figure 8 is the fused image obtained in the embodiment of the present invention.

[0042] Fig. 9 It is a flow chart of the in-situ imaging method of plankton according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The present invention is further described below with reference to the accompanying drawings and in combination with preferred embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0044] It should be noted that the directional terms such as left, right, up, down, top, and bottom in this embodiment are merely relative concepts or are based on the normal use status of the product, and should not be considered as restrictive.

[0045] The inventors have found that the detection methods in the prior art lack information on smaller-sized micro-plankton (5μm to 50μm) and biological components of background particles, and thus cannot achieve effective observation. Therefore, an embodiment of the present invention proposes an in-situ imaging system for plankton used in open scenes, which is based on a dual-telecentric microscopic back imaging optical path and simultaneously performs chlorophyll fluorescence imaging on plankton targets within the field of view, thereby achieving simultaneous back imaging and fluorescence imaging. The back imaging method can provide clear morphological features and high-contrast edges; chlorophyll fluorescence imaging can supplement classification information for targets with insufficient morphological information, and distinguish between biological and non-biological components in particles. The fluorescence image is fused with the back image after motion blur discrimination and processing, and the resulting fused image increases the information dimension, and can supplement chlorophyll information on the basis of morphological information, thereby improving target detection performance and facilitating the extraction of biological components of background particles.

[0046] like Figure 1 As shown, the plankton in-situ imaging system applied in an open scene according to an embodiment of the present invention comprises: a power supply, a telecentric light source 2, a dichroic beam splitter 5, a back imaging module 6, a fluorescence imaging module 7, a control module 8 and an image processing module 9, wherein the back imaging module 6 and the fluorescence imaging module 7 are located in a sealed cabin 4. The back imaging module 6 comprises a first telecentric lens 10 and a CMOS camera 11, and the back imaging module 6 uses the first telecentric lens 10 to receive the irradiation light reflected by the dichroic beam splitter 5, and performs back imaging on the CMOS camera 11; the fluorescence imaging module 7 comprises a filter 12, a second telecentric lens 13 and an sCMOS camera 14, and the fluorescence imaging module 7 uses the filter 12 to filter out wavelength components other than fluorescence, and performs fluorescence imaging on the sCMOS camera 14 through the second telecentric lens 13. The power supply 1 is used to power the telecentric light source 1, the CMOS camera 11, the sCMOS camera 14 and the control module 8, and the control module 8 is connected to the CMOS camera 11, the sCMOS camera 14 and the image processing module 9 respectively.

[0047] like Figure 2As shown, a telecentric light source 2 is incident on a plankton plane in a water body, and a plankton target 3 containing chlorophyll emits a longer wavelength fluorescence component under the excitation of the light (blue light) emitted by the telecentric light source 2, which is mixed with the illumination light, and the two are separated by a dichroic beam splitter; the back imaging module 6 performs back imaging on the separated illumination light, and the fluorescence imaging module 7 performs fluorescence imaging on the separated fluorescence; the control module 8 controls the back imaging module 6 and the fluorescence imaging module 7 to be triggered simultaneously, so as to realize the same frequency sampling of the backlight image and the fluorescence image of the same field of view.

[0048] The imaging principle of this embodiment is as follows:

[0049] The back imaging optical path uses an illumination light source to emit a uniform field of view from the back of the target, which is received by the camera on the other side to obtain clear edges and high-contrast contours. Since most populations of plankton have high transparency, the back imaging method combined with a telecentric illumination and telecentric lens dual telecentric system can accurately obtain its size, edge information and other morphological features.

[0050] Fluorescence is a photoluminescent phenomenon. When a fluorescent molecule is excited, its extranuclear electrons absorb energy and transition from the ground state to the excited state. The excited electrons have high energy and are unstable. They transition back to the ground state through radiation and release energy in the form of photons, and emit light with a longer wavelength than the incident light (usually in the visible light band), which is fluorescence. Chlorophyll is commonly found in plankton, especially in algae and other phytoplankton. Chlorophyll fluorescence imaging can obtain the distribution information of chlorophyll contained in plankton.

[0051] Chlorella, which is common in coastal waters, was selected as a typical microalgae for the experiment. Figures 3a to 3b As shown in the figure, according to the fluorescence excitation spectrum of Chlorella measured by the microplate reader, it can be seen that Chlorella has a higher light energy absorption efficiency in the blue light band, which can make the chlorophyll molecules in a higher excited state and obtain a higher fluorescence intensity value; under the condition of a fixed excitation light wavelength, the emission spectrum of Chlorella is measured by the microplate reader, and the peak of the emission spectrum curve corresponds to a wavelength of about 680nm. Therefore, in the fluorescence imaging module, a 460nm LED telecentric light source is selected as the illumination light source and a high-efficiency fluorescence excitation light source, and a long-wave filter is placed in front of the fluorescence receiving end to filter out the 460nm short-wavelength illumination light and retain the fluorescence in the long-wave band of about 680nm.

[0052] Specifically, in the imaging system of this embodiment, the 460nm telecentric light source 2 penetrates the sealed cabin 4 to irradiate the phytoplankton plane in the water body with uniform illumination, and the chlorophyll-containing phytoplankton target 3 emits a longer wavelength fluorescent component under the excitation of the excitation light (blue light). Figure 1The solid arrows in the figure show the fluorescence emitted by the plankton through transmission and scattering, and the longer wavelength fluorescence emitted by the plankton (as shown in the figure). Figure 1 The two parts are overlapped together (as shown by the dotted arrows in the figure). In order to separate the two parts, a dichroic beam splitter 5 is placed at the entrance of the sealed cabin 4, at a 45-degree angle to the incident illumination light and fluorescence. The fluorescence is directly transmitted and filtered out by the filter 12 (high cut-off depth filter), while the illumination light is reflected and received by the CMOS (Complementary Metal Oxide Semiconductor) camera 11 for back imaging. The fluorescence imaging module 7 uses the filter 12 (high cut-off depth filter) to filter out the wavelength components other than the fluorescence, and performs fluorescence imaging on a highly sensitive scientific research-level CMOS (sCMOS) camera through the second telecentric lens 13 with a 4-fold magnification; and the back imaging module 6 uses the first telecentric lens 10 with a uniform specification of 4-fold magnification to receive the illumination light reflected by the dichroic beam splitter 5, and performs back imaging on the CMOS camera 11. The control module 8 controls the CMOS camera 11 and the sCMOS camera 14 to trigger at the same time, so as to achieve the same frequency sampling of the backlight image and the fluorescence image of the same field of view. The collected background image and fluorescence image are transmitted to the image processing module 9 for processing.

[0053] Furthermore, this embodiment establishes a fluorescence illumination model to determine the illumination of the selected telecentric light source, the working distance of the telecentric light source and the telecentric lens, under the premise of ensuring the image plane illumination of the fluorescence signal received by the sCMOS camera 14 under certain chlorophyll concentration conditions. The process of establishing the fluorescence illumination model includes: the process of the excitation light propagating to the target object, the fluorescence process of chlorophyll, and the process of the fluorescence emission light being received by the camera. The details are as follows:

[0054] Seawater has a strong absorption and scattering effect on the propagation of light. For monochromatic parallel light, the illumination propagating in water can be described by exponential decay:

[0055] E(λ,r)==E0e -α(λ)r

[0056] Among them, E(λ,r) is the illumination of a monochromatic light source with a wavelength of λ transmitted to a point r away from the reference point; E0 is the illumination of the initial reference point, which is a constant; e is a natural constant, and α(λ) is the attenuation coefficient, which is related to the wavelength of light and can be divided into two parts: absorption coefficient and scattering coefficient, namely:

[0057] α(λ)=a(λ)+b(λ)

[0058] a(λ) is the absorption coefficient and b(λ) is the scattering coefficient.

[0059] Therefore, for the excitation light process, we can approximate the wavelength to be λ0 and the initial illumination to be Elight source When the excitation light propagates to the target position at a distance r1 from the light source, the illumination magnitude E1 is:

[0060]

[0061] a(λ0) and b(λ0) are the absorption coefficient and scattering coefficient of light with wavelength λ0 underwater respectively.

[0062] For the fluorescence process of chlorophyll, an efficiency factor κ is used to describe the efficiency of absorbing excitation light and emitting fluorescence, which is related to the wavelength of the excitation light; in addition, the fluorescence illumination size E emitted by the target plane emission (λ0,r1) is positively correlated with the chlorophyll concentration c:

[0063] E emission (λ0,r1)=κ(λ0)cE1(λ0,r1)

[0064] For the process of fluorescent emission light being received by the camera, the illumination received by the camera plane at a distance r2 from the target is:

[0065]

[0066] where λ e is the wavelength of the fluorescent light emitted.

[0067] The collected fluorescence image and background image are transmitted to the image processing module 9 for image processing. The image processing process is as follows: Figure 4 As shown. Since the illumination of chlorophyll fluorescence is low and the exposure time of the fluorescence receiving end is long, when the moving speed of the moving target in the water body is fast, the image will produce motion blur. In order to realize the determination of target motion blur, this embodiment establishes a scene classification data set, which includes three types of images: clear image, defocused blurred image, and motion blurred image. Based on the lightweight network (ShuffleNet), the scene classification model is trained using the above scene classification data set to obtain a scene classification model, which can determine whether the fluorescence image has motion blur in an end-to-end manner, and perform image processing to eliminate the blur for the image determined to be motion blurred. Since there is no movement at the camera end during the exposure time, and the movement of plankton can be approximately regarded as uniform linear motion, a more mature method for eliminating linear motion blur is adopted, and the motion blur kernel is solved for deconvolution to eliminate motion blur. Then, the fluorescence image and the background image are subjected to image fusion operation, and the weighted average method is used to perform pixel-level fusion of the fluorescence image and the background image in the spatial domain to obtain a fused image containing chlorophyll fluorescence information and sharp edges, which lays the foundation for image-based target detection of plankton.

[0068] The weighted average fusion method is to make a weighted average of each pixel of the original image. Taking the fluorescent image M and the background image N as an example, the pixel value of the fused image G is obtained by making a weighted average of the pixel values ​​of the fluorescent image M and the background image N at the corresponding position, that is:

[0069] G(i,j)=w i M(i,j)+w2N(i,j)

[0070] Where G(i,j) is the pixel value at position (i,j) of the fused image, M(i,j) is the pixel value at position (i,j) of the fluorescence image, and N(i,j) is the pixel value at position (i,j) of the background image; w1 and w2 are weighting coefficients, and satisfy w1+w2=1.

[0071] In this embodiment, the original background grayscale image is converted into an RGB three-channel color image, and in the R channel, the fluorescence image pixel value is weightedly fused with the background image, so that the chlorophyll-containing part 15 appears red. Figures 5 to 8 As shown in the figure, respectively, are the fluorescence image with motion blur, the background image, the fluorescence image after eliminating the motion blur, and the obtained fused image collected in this embodiment.

[0072] like Fig. 9 As shown, an embodiment of the present invention also provides a method for in-situ imaging of plankton in an open scene, comprising the following steps:

[0073] S1. A telecentric light source is passed through a sealed cabin to illuminate the phytoplankton plane in the water body with uniform illumination. The chlorophyll-containing phytoplankton targets emit longer wavelength fluorescence components under the light excitation of the telecentric light source; the illumination light is transmitted and scattered by the phytoplankton and is mixed with the excited fluorescence.

[0074] S2. Separate the mixed illumination light and fluorescence.

[0075] S3. Perform back imaging on the separated illumination light and perform fluorescence imaging on the separated fluorescence at the same time.

[0076] S4. Perform the same frequency sampling on the backlight image and the fluorescence image of the same field of view.

[0077] In some embodiments, the step is further included: S5, performing motion blur identification and processing on the fluorescent image, and then fusing it with the background image to obtain a fused image.

[0078] Specifically, step S5 includes:

[0079] S51. Establish a scene classification data set, including: clear images, out-of-focus blurred images, and motion blurred images.

[0080] S52. Based on the lightweight network, the scene classification data set is used for training to obtain a scene classification model, and whether there is motion blur in the fluorescent image is determined in an end-to-end manner.

[0081] S53. If there is no motion blur in the fluorescence image, directly proceed to step S44; if there is motion blur in the fluorescence image, solve the motion blur kernel and perform deconvolution to eliminate the motion blur.

[0082] S54, using a weighted average method to perform pixel-level fusion of the fluorescence image and the background image in the spatial domain, to obtain a fused image that finally contains chlorophyll fluorescence information and sharp edges.

[0083] Furthermore, step S54 specifically includes: performing weighted averaging on each pixel of the fluorescent image M and the background image N, and the pixel value of the fused image G is obtained by performing weighted averaging on the pixel values ​​of M and N at corresponding positions, that is:

[0084] G(i,j)=w1M(i,j)+w2N(i,j)

[0085] Where G(i,j) is the pixel value at position (i,j) of the fused image, M(i,j) is the pixel value at position (i,j) of the fluorescence image, and N(i,j) is the pixel value at position (i,j) of the background image; w1 and w2 are weighting coefficients, and satisfy w1+w2=1;

[0086] The original background grayscale image is converted into an RGB three-channel color image, and in the R channel, the pixel values ​​of the fluorescence image are weightedly fused with the background image, so that the part containing chlorophyll appears red.

[0087] The in-situ imaging system and method for plankton applied in open scenes proposed in the embodiments of the present invention can perform in-situ sampling in nearshore waters, and the obtained fusion image of background and fluorescence can solve the problems of biological information discrimination in micro-plankton particles and accurate classification and identification of plankton, thereby realizing effective observation of micro-plankton.

[0088] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art of the present invention, several equivalent substitutions or obvious variations can be made without departing from the concept of the present invention, and the performance or use is the same, which should be regarded as belonging to the protection scope of the present invention.

Claims

1. An in-situ imaging system for plankton used in an open scene, characterized in that: include: Telecentric light source, dichroic beam splitter, background imaging module, fluorescence imaging module, sealed cabin, control module and image processing module; A telecentric light source is incident on a plankton plane in a water body, and plankton targets containing chlorophyll emit longer wavelength fluorescence components under the excitation of the light emitted by the telecentric light source, which are mixed with the illumination light, and the dichroic beam splitter separates the two; the back imaging module performs back imaging on the separated illumination light, and the fluorescence imaging module performs fluorescence imaging on the separated fluorescence; the control module controls the back imaging module and the fluorescence imaging module to be triggered simultaneously, so as to achieve the same frequency sampling of the backlight image and the fluorescence image of the same field of view; the image processing module is also used for motion blur discrimination and processing of the collected fluorescence image, and then fusing it with the back image to obtain a fused image; the back imaging module and the fluorescence imaging module include telecentric lenses, which form a dual telecentric system in conjunction with telecentric illumination.

2. The plankton in-situ imaging system for use in an open scene as claimed in claim 1, characterized in that: The back image imaging module comprises a first telecentric lens and a CMOS camera. The back image imaging module uses the first telecentric lens to receive the illumination light reflected by the dichroic beam splitter and performs back image imaging on the CMOS camera.

3. The plankton in-situ imaging system for use in an open scene as claimed in claim 1, characterized in that: The fluorescence imaging module comprises a filter, a second telecentric lens and a sCMOS camera. The fluorescence imaging module uses the filter to filter out wavelength components other than fluorescence, and performs fluorescence imaging on the sCMOS camera through the second telecentric lens.

4. The plankton in-situ imaging system for use in an open scene as claimed in claim 1, characterized in that: A scene classification dataset is established through the image processing module, and a scene classification model is obtained by training with the scene classification dataset based on a lightweight network. Whether the fluorescence image has motion blur is determined in an end-to-end manner. For the fluorescence image with motion blur, the motion blur kernel is solved for deconvolution to eliminate the motion blur. The fluorescence image and the background image are fused at the pixel level in the spatial domain using the weighted averaging method to obtain a fused image containing chlorophyll fluorescence information and sharp edges.

5. The plankton in-situ imaging system for use in an open scene as claimed in claim 1, characterized in that: By establishing a fluorescence illumination model, the illumination of the telecentric light source, the working distance of the telecentric light source and the telecentric lens are determined under the premise that the illumination of the image plane where the fluorescence imaging module receives the fluorescence signal can be guaranteed under certain chlorophyll concentration conditions.

6. The plankton in-situ imaging system for use in an open scene as claimed in claim 4, characterized in that: The process of establishing the fluorescence illumination model includes: the process of the excitation light propagating to the target object, the fluorescence process of chlorophyll, and the process of the fluorescence emission light being received by the fluorescence imaging module.

7. The plankton in-situ imaging system for use in an open scene as claimed in claim 1, characterized in that: The back imaging module and the fluorescence imaging module are located in the sealed cabin; the dichroic beam splitter is placed at the entrance of the sealed cabin, forming a 45-degree angle with the incident illumination light and fluorescence.

8. The plankton in-situ imaging system for use in an open scene as claimed in claim 1, characterized in that: The telecentric light source is a 460nm telecentric light source, which penetrates the sealed cabin and impinges on the plankton plane in the water body with uniform illumination.

9. A method for in-situ imaging of plankton in an open scene based on the system of claim 1, characterized in that: The following steps are involved: S1. A telecentric light source is passed through a sealed cabin to illuminate the phytoplankton plane in the water body with uniform illumination. The chlorophyll-containing phytoplankton targets emit longer wavelength fluorescence components under the light excitation of the telecentric light source; the illumination light is transmitted and scattered by the phytoplankton and is mixed with the excited fluorescence; S2, separating the overlapping illumination light and fluorescence; S3, performing back imaging on the separated illumination light, and performing fluorescence imaging on the separated fluorescence; S4. Perform the same frequency sampling on the backlight image and the fluorescence image of the same field of view.

10. The method for in-situ imaging of plankton in an open scene as claimed in claim 9, characterized in that: The method further includes the step: S5, performing motion blur identification and processing on the fluorescent image, and then fusing the fluorescent image with the background image to obtain a fused image.

11. The method for in-situ imaging of plankton in an open scene according to claim 10, characterized in that: Step S5 includes: S51, establishing a scene classification data set, including: clear images, out-of-focus blurred images, and motion blurred images; S52, training a scene classification model using the scene classification data set based on a lightweight network, and determining whether there is motion blur in the fluorescence image in an end-to-end manner; S53, if there is no motion blur in the fluorescent image, directly proceed to step S44; if there is motion blur in the fluorescent image, solve the motion blur kernel and perform deconvolution to eliminate the motion blur; S54, using a weighted average method to perform pixel-level fusion of the fluorescence image and the background image in the spatial domain, to obtain a fused image that finally contains chlorophyll fluorescence information and sharp edges.

12. The method for in-situ imaging of plankton in an open scene according to claim 11, characterized in that: Step S54 specifically includes: taking a weighted average of each pixel of the fluorescent image M and the background image N, and the pixel value of the fused image G is obtained by taking a weighted average of the pixel values ​​of M and N at corresponding positions, that is: ； in To fuse the images The pixel value at the position, For fluorescence images The pixel value at the position, For back image The pixel value at the position; and is the weighting coefficient and satisfies ; The original background grayscale image is converted into an RGB three-channel color image, and in the R channel, the pixel values ​​of the fluorescence image are weightedly fused with the background image, so that the part containing chlorophyll appears red.

Images