Method for automatically measuring biomass of freshwater zooplankton based on binocular vision

Abstract

The invention discloses a zooplankter biomass automatic measurement method based on binocular vision, and belongs to the field of resources and environmental. The zooplankter type and an external rectangular frame are obtained by using a Yolo target detection algorithm, and morphological parameters of the body length and the body width of floating movement are obtained by using a morphological digital image processing technology; according to the method, the body thickness parameter of the zooplankton target area is calculated by adopting the visual difference of left and right microscopic images of the binocular microscope, and finally the biomass of the zooplankton is calculated by substituting the body length, the body width and the body thickness into the approximate quadrature formula of the corresponding category, so that automatic measurement of the biomass of the zooplankton is realized, and the measurement efficiency of existing manual analysis is effectively improved.

Description

Technical Field

[0001] The invention belongs to the field of resources and environment, and in particular relates to a method for automatically measuring the biomass of freshwater zooplankton based on binocular vision. Background Art

[0002] Zooplankton biomass is a key indicator for monitoring aquatic ecosystems. Currently, the primary measurement method is volumetric analysis, which uses a relative density of 1 and measures zooplankton volume to determine biomass. Common zooplankton volumetric methods include sedimentation volumetric analysis, drainage volumetric analysis, and morphometric volumetric analysis. Sedimentation volumetric analysis involves fixing a sample for 24 hours and estimating its volume by reading the sediment, while drainage volumetric analysis involves filtering the sample into a quantitative graduated cylinder and estimating its volume by reading the water level rise mark. Both methods suffer from low accuracy. The morphometric volumetric analysis uses the morphological parameters of zooplankton (length, width, and thickness) from microscopic images to calculate zooplankton biomass (for an approximate volumetric formula, refer to the "Technical Requirements for Water Ecological Monitoring: Freshwater Zooplankton" (Trial Version), issued by the China National Environmental Monitoring Center in 2022). This is the most commonly used biomass measurement method in aquatic ecosystem surveys. However, the measurement process requires manual microscopic analysis of morphological parameters such as length, width, and thickness, making it time-consuming and labor-intensive, and unable to meet the technical requirements of large-scale, high-frequency aquatic biomonitoring. The present invention discloses an automatic measurement method for zooplankton biomass based on binocular vision. The method comprises the following steps: using a binocular microscope to collect left and right microscopic images of zooplankton and perform stereo matching to obtain a depth map; using a Yolo target detection algorithm to obtain the zooplankton category and a bounding box; and adopting digital morphological image processing technology to obtain the body length and body width morphological parameters of the zooplankton; using the visual difference between the left and right microscopic images of the binocular microscope to calculate the body thickness parameter of the zooplankton target area; finally, substituting the body length, body width and body thickness into the approximate quadrature formula corresponding to the zooplankton category to calculate the zooplankton biomass, thereby realizing automatic measurement of the zooplankton biomass and effectively improving the measurement efficiency of the existing manual analysis. Summary of the Invention

[0003] The existing method for measuring zooplankton biomass is to first manually measure the morphological parameters of zooplankton such as body length, body width and body thickness under a microscope, and then input the morphological parameters into an approximate geometric figure to obtain the volume of zooplankton, thereby calculating the zooplankton biomass. At present, the measurement of zooplankton biomass requires a large amount of manpower and labor, and cannot meet the technical needs of large-scale and high-frequency aquatic biological monitoring. The present invention discloses an automatic measurement method for zooplankton biomass based on binocular vision, which uses the Yolo target detection algorithm to obtain the zooplankton category and bounding box, and uses digital morphological image processing technology to obtain the body length and body width morphological parameters of plankton; uses the visual difference between the left and right microscopic images of a binocular microscope to calculate the body thickness parameters of the zooplankton target area, and finally substitutes the body length, body width and body thickness into the approximate quadrature formula corresponding to the zooplankton category to calculate the zooplankton biomass, thereby realizing the automatic measurement of zooplankton biomass, effectively improving the measurement efficiency of the existing manual analysis.

[0004] The technical solution of the present invention is as follows:

[0005] The automatic measurement method of zooplankton biomass based on binocular vision is implemented as follows:

[0006] Step 1: Binocular Microscope Image Acquisition

[0007] Place the zooplankton slide sample on the stage, and then use a computer program to drive the binocular vision camera at the left and right eyepieces of the binocular microscope to synchronously collect the left and right microscopic images of the zooplankton, thereby obtaining the left microscopic image I of the zooplankton in the same microscopic field of view. l and right micrograph I f .

[0008] Step 2: Stereo Matching

[0009] The local stereo matching algorithm is used to perform stereo matching on the left and right eye microscopic images to obtain the disparity map d of the zooplankton binocular microscopic image, and the depth map Z is obtained based on the principle of similar triangles:

[0010] ,

[0011] Where f is the focal length and b is the baseline value of the binocular camera.

[0012] Step 3: Image Recognition and Segmentation

[0013] The Yolo target detection method is used to detect the microscopic images of zooplankton such as rotifers to obtain the category of zooplankton and the external rectangular frame of the target area. The rectangular frame is mapped to the left microscopic image to obtain the segmentation result of zooplankton I s , using the Great Law ( ) to I sThe image is binarized to obtain the segmentation mask image M0.

[0014] ,

[0015] In the segmentation mask image, there are limbs of zooplankton such as feet, and the biomass of zooplankton is calculated by the volume of the trunk. The present invention removes the limbs such as feet through opening operation to obtain a binary image M1 that only retains the trunk of the zooplankton.

[0016] ,

[0017] in, is a morphological corrosion operation, is the morphological dilation operation, It is a structural element.

[0018] Step 4: Morphological parameter extraction

[0019] In the zooplankton microscopic image, compared with the relatively flat background area, the zooplankton target area has obvious spatial three-dimensional features, and the depth map Z can represent the spatial three-dimensional features of the microscopic image, that is, the value of each position in the depth map is the distance from the point to the camera plane. The left microscopic image I can be obtained by filling the background of the trunk binary image M1. l The segmented binary image M is obtained. The area with a value of 0 in the binary image M is the background area of ​​the microscopic image, and the area with a value of 1 is the zooplankton target area. The present invention calculates the zooplankton thickness by the difference between the depth value of the target area and the average depth value of the background area, that is:

[0020] ,

[0021] ,

[0022] in, is the average depth value of the background area, in the matrix In the image, each value of the target area is the spatial pixel distance corresponding to the location, and the maximum spatial distance is the body thickness of the zooplankton.

[0023] ,

[0024] in, is the pixel size of the binocular camera, is the magnification of the microscope.

[0025] To extract the length and width parameters of zooplankton, the present invention obtains the minimum rotation circumscribed rectangular box of the binary M1 by exhaustively enumerating the rotation angles. min =(l p ,w p ,ap ), where l p ,w p ,a p are the length, width and rotation angle of the rotating circumscribed rectangular frame, then the body length of the zooplankton and body width It can be obtained by the following formula:

[0026] ,

[0027] .

[0028] Step 5: Biomass calculation

[0029] According to the zooplankton category c identified in step (3), the morphological parameters of body length, body width and body thickness are substituted into the approximate quadrature formula f to calculate the biomass of zooplankton.

[0030] ,

[0031] in, is the relative density, usually taken as 1, The approximate quadrature formula for category C defined in the Technical Requirements for Water Ecological Monitoring of Freshwater Zooplankton (Trial) is: is the biomass of zooplankton.

[0032] Beneficial effects:

[0033] The present invention discloses an automatic measurement method for zooplankton biomass based on binocular vision. The zooplankton area is detected by Yolo, the body length and thickness of the zooplankton trunk are obtained using morphological digital image processing technology, and the body thickness parameter of the zooplankton is obtained using the depth map of the binocular microscopic image. Zooplankton biomass is measured in combination with the zooplankton approximate volume calculation formula defined in the "Technical Requirements for Freshwater Zooplankton in Water Ecological Monitoring" (Trial). BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 : Implementation steps of an automatic zooplankton biomass measurement method based on binocular vision.

[0035] Figure 2 :Example diagram of biomass calculation of Brachionus calyciflorus rotifer (I).

[0036] Figure 3 :Example diagram of biomass calculation of Brachionus calyciflorus rotifer (2).

[0037] Figure 4 :Example diagram of biomass calculation of long-limbed rotifer (a).

[0038] Figure 5:Example diagram of biomass calculation of long-trilimbed rotifer (2). DETAILED DESCRIPTION

[0039] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. However, the following embodiments are intended only to explain the present invention, and the scope of protection of the present invention should include the entire contents of the claims. Moreover, through the description of the following embodiments, those skilled in the art can fully implement the entire contents of the claims of the present invention.

[0040] Example

[0041] The automatic measurement method of zooplankton biomass based on binocular vision is implemented as follows:

[0042] (1) Binocular microscopic image acquisition

[0043] Place the zooplankton slide sample on the stage, and then use a computer program to drive the binocular vision camera at the left and right eyepieces of the binocular microscope to synchronously collect the left and right microscopic images of the zooplankton, thereby obtaining the left microscopic image I of the zooplankton in the same microscopic field of view. l and right micrograph I f .

[0044] (2) Stereo matching

[0045] The local stereo matching algorithm is used to perform stereo matching on the left and right eye microscopic images to obtain the disparity map d of the zooplankton binocular microscopic image, and the depth map Z is obtained based on the principle of similar triangles:

[0046] ,

[0047] Where f is the focal length and b is the baseline value of the binocular camera.

[0048] (3) Image recognition and segmentation

[0049] The Yolo target detection method is used to detect the microscopic images of zooplankton such as rotifers to obtain the category of zooplankton and the external rectangular frame of the target area. The rectangular frame is mapped to the left microscopic image to obtain the segmentation result of zooplankton I s , using the Great Law ( ) to I s The image is binarized to obtain the segmentation mask image M0.

[0050] ,

[0051] In the segmentation mask image, there are limbs of zooplankton such as feet, and the biomass of zooplankton is calculated by the volume of the trunk. The present invention removes the limbs such as feet through opening operation to obtain a binary image M1 that only retains the trunk of the zooplankton.

[0052] ,

[0053] in, is a morphological corrosion operation, is the morphological dilation operation, It is a structural element.

[0054] (4) Morphological parameter extraction

[0055] In the zooplankton microscopic image, compared with the relatively flat background area, the zooplankton target area has obvious spatial three-dimensional features, and the depth map Z can represent the spatial three-dimensional features of the microscopic image, that is, the value of each position in the depth map is the distance from the point to the camera plane. The left microscopic image I can be obtained by filling the background of the trunk binary image M1. l The segmented binary image M is obtained. The area with a value of 0 in the binary image M is the background area of ​​the microscopic image, and the area with a value of 1 is the zooplankton target area. The present invention calculates the zooplankton thickness by the difference between the depth value of the target area and the average depth value of the background area, that is:

[0056] ,

[0057] ,

[0058] in, is the average depth value of the background area, in the matrix In the target area, each value is the spatial pixel distance corresponding to the location, and the maximum spatial distance is the body thickness of the zooplankton.

[0059] ,

[0060] in, is the pixel size of the binocular camera, is the magnification of the microscope.

[0061] To extract the length and width parameters of zooplankton, the present invention obtains the minimum rotation circumscribed rectangular box of the binary M1 by exhaustively enumerating the rotation angles. min =(l p ,w p ,a p ), where l p ,w p ,a p are the length, width and rotation angle of the rotating circumscribed rectangular frame, then the body length of the zooplankton and body width It can be obtained by the following formula:

[0062] ,

[0063] .

[0064] (5) Biomass calculation

[0065] According to the zooplankton category c identified in step (3), the morphological parameters of body length, body width and body thickness are substituted into the approximate quadrature formula f to calculate the biomass of zooplankton.

[0066] ,

[0067] in, is the relative density, usually taken as 1, It is the approximate quadrature formula for category C defined in the Technical Requirements for Water Ecological Monitoring of Freshwater Zooplankton (Trial). is the biomass of zooplankton.

[0068] The foregoing is merely a list of specific embodiments of the present application, intended to enable those skilled in the art to understand or implement the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the broadest scope consistent with the principles and novel features of the present application.

Claims

1. The method for automatically measuring zooplankton biomass based on binocular vision is characterized in that: The steps include: Step 1: binocular microscope image acquisition; Step 2: Stereo matching; Step 3: Image recognition and segmentation; Step 4: Extract morphological parameters; Step 5: Biomass calculation.

2. The method for automatically measuring zooplankton biomass based on binocular vision according to claim 1, characterized in that: In step 1, a slide sample of zooplankton is placed on the stage, and then a computer program is used to drive a binocular vision camera at the left and right eyepieces of a binocular microscope to synchronously collect left and right eye microscopic images of zooplankton, thereby obtaining a left microscopic image I of zooplankton in the same microscopic field of view. l and right micrograph I f .

3. The method for automatically measuring zooplankton biomass based on binocular vision according to claim 1, characterized in that: In step 2, the left and right eye microscopic images are stereo matched using a local stereo matching algorithm to obtain a disparity map d of the zooplankton binocular microscopic image, and a depth map Z is obtained based on the principle of similar triangles.

4. The method for automatically measuring zooplankton biomass based on binocular vision according to claim 3 is characterized in that: In step 2, , where f is the focal length and b is the baseline value of the binocular camera.

5. The method for automatic measurement of zooplankton biomass based on binocular vision according to claim 1, characterized in that: In step 3, the Yolo target detection method is used to detect the microscopic images of zooplankton such as rotifers to obtain the category of zooplankton and the external rectangular frame of the target area, and the rectangular frame is mapped to the left microscopic image to obtain the segmentation result of zooplankton I s , using the Great Law ( ) to I s The image is binarized to obtain a segmentation mask image M0, and the limbs such as the feet are removed through an opening operation to obtain a binary image M1 that only retains the trunk of the zooplankton.

6. The method for automatically measuring zooplankton biomass based on binocular vision according to claim 5, characterized in that: In step three, ; ,in, is a morphological corrosion operation, is the morphological dilation operation, It is a structural element.

7. The method for automatically measuring zooplankton biomass based on binocular vision according to claim 1, characterized in that: In step 4, the trunk binary image M1 is background filled to obtain the left microscopic image I l The segmented binary image M is obtained. The area with a value of 0 in the binary image M is the background area of ​​the microscopic image, and the area with a value of 1 is the target area of ​​the zooplankton. The thickness of the zooplankton is calculated by the difference between the depth value of the target area and the average depth value of the background area, and the body length and width parameters of the zooplankton are extracted.

8. The method for automatically measuring zooplankton biomass based on binocular vision according to claim 7, characterized in that: In step four: ， ， in, is the average depth value of the background area, in the matrix In the target area, each value is the spatial pixel distance corresponding to the position, and the maximum spatial distance is the body thickness of the zooplankton. ， in, is the pixel size of the binocular camera, is the magnification of the microscope; By exhaustively enumerating the rotation angles, we can obtain the minimum rotational bounding rectangle box of the binary M1. min =(l p ,w p ,a p ), where l p 、w p 、a p are the length, width and rotation angle of the rotating circumscribed rectangular frame, the body length of the zooplankton and body width Obtained by the following formula: ， 。 9. The method for automatically measuring zooplankton biomass based on binocular vision according to claim 1, characterized in that: In step five, according to the zooplankton category c identified in step three, the morphological parameters of body length, body width and body thickness are substituted into the approximate quadrature formula f to calculate the biomass of zooplankton.

10. The method for automatic measurement of zooplankton biomass based on binocular vision according to claim 9, characterized in that: In step five, ,in, is the relative density, which takes the value of 1. This is the approximate quadrature formula for category C defined in the Technical Requirements for Water Ecological Monitoring of Freshwater Zooplankton (Trial), is the biomass of zooplankton.

Images