Zebra fish growth and development detection method, system and equipment and storage medium

Abstract

The invention provides a zebra fish growth and development detection method, system and equipment and a storage medium. The method comprises the following steps: calibrating a zero point position of a to-be-detected target hole site of a to-be-detected porous cell culture plate; controlling a forward camera to collect forward images of the to-be-detected target hole sites until the forward images of all the to-be-detected target hole sites in the to-be-detected porous cell culture plate are collected; controlling an inverted camera to collect inverted images of the to-be-detected target hole sites until the inverted images of all the to-be-detected target hole sites in the to-be-detected porous cell culture plate are collected; calculating the ratio of zebra fish tissues in the positive image and the negative image corresponding to each target hole site to be detected, and selecting the positive image or the negative image with the large ratio as an image to be detected; and by utilizing the trained zebra fish growth and development detection model, based on the to-be-detected image, obtaining growth and development classification information of the zebra fish in the to-be-detected target hole site. The problems of low efficiency, large error and complicated operation in manual observation of zebra fish egg growth and development are effectively solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of biological experiment detection, and particularly relates to a method, a system, a device and a storage medium for detecting the growth and development of zebrafish. Background Art

[0002] In the field of biological experiments, fertilized eggs and embryos of zebrafish are commonly used experimental materials, which are widely used in pharmacological research, water quality monitoring and other aspects. Taking the experiment of measuring the acute toxicity of water quality by using the zebrafish egg method as an example, it is necessary to judge the experimental end point based on indicators such as oocyte condensation, somite formation, whether the tail is separated, and the presence or absence of heartbeat, and then analyze the water quality status. However, currently, it mainly relies on manual observation of the growth and development of zebrafish embryos in a multi-well cell culture plate under an optical microscope, and this method has many drawbacks.

[0003] The efficiency of manually identifying zebrafish microscope images is extremely low. When dealing with a large number of samples, the problem of slow speed is particularly prominent, and in order to ensure the observation effect, it is necessary to additionally fix the position of fish eggs or larvae. According to national standards, in the experiment of measuring water quality with different concentrations of zebrafish, it is necessary to observe the growth and development status of zebrafish eggs at 24h and 48h respectively. This not only consumes a large amount of time and manpower, but also is easily affected by subjective factors such as observer fatigue, experience, and inconsistent judgment criteria, resulting in errors in the observation results. In addition, when zebrafish move in a multi-well cell culture plate, the microscope field of view is small, and it is extremely time-consuming to manually locate the position of zebrafish and switch samples for observation, seriously affecting the experimental efficiency and accuracy. Summary of the Invention

[0004] The purpose of the present application is to provide a method, a system, a device and a storage medium for detecting the growth and development of zebrafish, so as to solve the problems of low efficiency, large error and cumbersome operation in the existing manual observation of the growth and development of zebrafish eggs, and realize the efficient, accurate and automatic detection of the growth and development status of zebrafish eggs.

[0005] According to one aspect of the present application, a method for detecting the growth and development of zebrafish is proposed, and the method includes the following steps:

[0006] S1. Obtain the multi-well cell culture plate to be tested within the field of view of the positioning camera, and calibrate the zero position of the target well to be tested on the multi-well cell culture plate to be tested;

[0007] S2. Control the upright camera to collect the upright images of the target wells to be tested until the collection of the upright images of all the target wells to be tested in the multi-well cell culture plate to be tested is completed;

[0008] After recalibrating the zero position of the target pore to be measured, control the inverted camera to collect the inverted images of the target pore to be measured until the inverted images of all the target pores to be measured in the multi-well cell culture plate to be measured are collected. Among them, the upright images and the inverted images contain the image information of zebrafish tissues;

[0009] S3. Calculate the occupation ratio of the zebrafish tissues in the upright image and the inverted image corresponding to each target pore to be measured, and select the upright image or the inverted image with a larger occupation ratio as the image to be measured;

[0010] S4. Using the trained zebrafish growth and development detection model, based on the image to be measured, obtain the growth and development classification information of the zebrafish in the target pore to be measured.

[0011] In the above technical solution, by first calibrating the zero position and then using the upright camera component and the inverted camera component to collect zebrafish images respectively, the image information of zebrafish can be obtained from different angles, providing a comprehensive data basis for accurately judging its growth and development classification information in the subsequent stage, and improving the accuracy and comprehensiveness of the detection.

[0012] Further, the steps of automatic focusing of the upright camera or the inverted camera in step S2 include:

[0013] S21. Continuously collect the original images of the target pore to be measured, and perform Laplace filtering on the original images;

[0014] S22. Calculate the gray variance of the image after Laplace filtering, traverse different focal length positions of the lens in the upright camera or the inverted camera, record the corresponding gray variance change curve, and determine the mechanical position corresponding to the maximum gray variance as the best focus point;

[0015] S23. Based on the best focus point, complete the collection of the upright image or the inverted image of the target pore to be measured.

[0016] In the above technical solution, by performing Laplace filtering on the continuously collected original images, calculating the gray variance and traversing different focal length positions of the lens to determine the best focus point, it can ensure that the upright camera and the inverted camera are accurately focused at each target pore to be measured, and then clear and accurate upright images and inverted images are collected, providing a high-quality data basis for accurately obtaining the growth and development classification information of zebrafish in the subsequent stage.

[0017] Further, the construction steps of the zebrafish growth and development detection model in step S4 include:

[0018] S41. Collect images of zebrafish development, label the normal state, abnormal state, and the corresponding development types of the normal and abnormal states in each image, and construct an image dataset containing zebrafish growth and development classification information. The images include top-view images and bottom-view images of zebrafish;

[0019] S42. Use the YOLOV8 object detection algorithm to construct a binary classification model for detecting zebrafish growth and development, and based on the image dataset, train and optimize the network parameters in the model to obtain a zebrafish growth and development detection model. Among them, the zebrafish growth and development detection model includes a detection head, an EfficientNet feature extractor, and a feature fusion module combined by a feature pyramid network and a path aggregation network.

[0020] In the above technical solution, by collecting rich zebrafish development images and labeling them, constructing a targeted image dataset, and then using the advanced YOLOV8 algorithm and specific feature extractors and feature fusion modules to construct a model, the model can learn the characteristic patterns of zebrafish growth and development, so as to accurately detect and classify the growth and development status of zebrafish.

[0021] Furthermore, the development types in the normal state include zygotic stage, cleavage stage, blastula stage, gastrula stage, somite stage, primordium stage, hatching stage, and larva. The development types in the abnormal state include egg coagulation, somite not formed, and tail not separated. Defining the specific development types in the normal and abnormal states makes the classification of the growth and development status of zebrafish more detailed and accurate, providing an accurate judgment basis for related research and applications.

[0022] Furthermore, step S41 includes removing abnormal images, enhancing images, and normalizing the collected images. The enhanced image processing includes performing translation, rotation, scaling, and mirroring operations on the top-view image and / or bottom-view image.

[0023] In the above technical solution, removing abnormal images can improve the quality of the dataset and reduce interference factors; enhancing the images can expand the dataset, increase the diversity of the data, and improve the generalization ability of the model, enabling it to more accurately detect the growth and development status of zebrafish in different scenarios.

[0024] Furthermore, step S42 includes;

[0025] S421. Construct the YOLOv8 model structure. The YOLOv8 model includes a detection head, an EfficientNet feature extractor, and a feature fusion module combined by a feature pyramid network and a path aggregation network. The detection head is used to generate prediction boxes of different classes and output category, bounding box regression, and confidence information;

[0026] S422. Perform model training and optimization. Train the hyperparameters of the YOLOv8 model based on the image dataset, and optimize the YOLOv8 model by calculating the loss function, adjusting the learning rate, and optimizing the hyperparameter combination. Among them, the hyperparameters include the number of training epochs, batch size, input image size, initial learning rate, and learning decay coefficient;

[0027] S423. Model evaluation and accuracy improvement. Evaluate the classification performance of the model by calculating the accuracy, precision, and F1 value until a zebrafish growth and development detection model that meets the preset requirements is obtained.

[0028] In a second aspect, the present application proposes a zebrafish growth and development detection system, which includes:

[0029] A zero-point position calibration module, configured to acquire a porous cell culture plate to be measured within the field of view of the positioning camera, and calibrate the zero-point position of the target hole to be measured on the porous cell culture plate to be measured;

[0030] An image acquisition device, configured to control the upright camera to acquire the upright images of the target holes to be measured until the upright images of all the target holes to be measured in the porous cell culture plate to be measured are acquired;

[0031] After re-calibrating the zero-point position of the target hole to be measured, control the inverted camera to acquire the inverted images of the target holes to be measured until the inverted images of all the target holes to be measured in the porous cell culture plate to be measured are acquired. Among them, the upright images and the inverted images contain the image information of the zebrafish tissue;

[0032] A to-be-measured image calculation module, configured to calculate the occupation ratio of the zebrafish tissue in the upright image and the inverted image corresponding to each target hole to be measured, and select the upright image or the inverted image with a larger occupation ratio as the to-be-measured image;

[0033] A zebrafish development type acquisition module, configured to use the trained zebrafish growth and development detection model to obtain the growth and development classification information of the zebrafish in the target hole to be measured based on the to-be-measured image.

[0034] Furthermore, the image acquisition device includes a support assembly and a positioning camera, an upright camera assembly, an inverted camera assembly, a control assembly, and a stage installed on the support assembly. The upright camera assembly and the inverted camera assembly are installed on the upper and lower sides of the stage through the control assembly, where:

[0035] The stage is provided with a transparent area, and the porous cell culture plate to be measured is placed in the transparent area;

[0036] The positioning camera is used to acquire the position of the porous cell culture plate to be measured and realize the zero-point position calibration of the target hole to be measured;

[0037] The upright camera assembly, including an upright light source and an upright camera, is used to capture the top-down image of zebrafish in the target hole to be measured;

[0038] The inverted camera assembly, including an inverted light source and an inverted camera, is used to capture the bottom-up image of zebrafish in the target hole to be measured;

[0039] The control assembly is used to move the upright camera assembly and the inverted camera assembly in a specific direction at a predetermined step size.

[0040] In the above technical solution, the structure of the image acquisition device is reasonably designed. By accurately calibrating the zero position with the positioning camera, the upright and inverted camera assemblies capture images from different angles, and the control assembly precisely controls the movement of the cameras, which can ensure the acquisition of high-quality and multi-angle zebrafish images, providing a reliable data source for subsequent detection.

[0041] Furthermore, the control assembly includes a first stepping motor, a second stepping motor, an X-axis track, and Y-axis tracks one and two located in the same plane. The upright camera assembly and the inverted camera assembly are respectively fixed on Y-axis tracks one and two. The first stepping motor is used to move the upright camera assembly and the inverted camera assembly along the Y-axis direction, and the second stepping motor is connected to the X-axis track to control the movement of the plane where Y-axis tracks one and two are located along the X-axis direction.

[0042] In the above technical solution, the structure design of the control assembly enables the upright camera assembly and the inverted camera assembly to move precisely in the X-axis and Y-axis directions, realizing accurate focusing and image acquisition of each target hole to be measured on the multi-well cell culture plate, improving the efficiency and accuracy of image acquisition.

[0043] Furthermore, the support assembly includes a first bracket, a second bracket, and an inverted "L"-shaped bracket for fixing the positioning camera. Y-axis tracks one and two are fixed on the X-axis track through the first bracket and the second bracket. The structure design of the support assembly provides a stable support for the entire image acquisition device, ensuring the stability of each component during operation, reducing the influence of factors such as vibration on the image acquisition quality, and guaranteeing the reliability of the detection system.

[0044] In a third aspect, the present application proposes a terminal device, including a processor, a memory, and a computer program stored in the memory. The computer program is executed by the processor to implement the zebrafish growth and development detection method as described in any one of the above.

[0045] In a fourth aspect, the present application proposes a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by the processor, the zebrafish growth and development detection method as described in any one of the above is implemented.

[0046] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0047] (1) Completeness of image acquisition: By using a dual-camera setup to capture the state images of zebrafish eggs from different angles, the problem of perspective occlusion caused by positioning is effectively avoided, and more comprehensive and accurate zebrafish egg image information can be obtained, providing a rich and reliable data basis for subsequent analysis and judgment.

[0048] (2) High degree of automation: Continuous manual monitoring is not required, and automatic image data acquisition can be achieved, greatly reducing labor costs and labor intensity. At the same time, the errors and instabilities that may be brought by manual operations are reduced, improving the efficiency and consistency of detection.

[0049] (3) Tightly integrating deep learning algorithms, machine vision technology, and cameras, giving full play to the advantages of each technology. Deep learning algorithms can accurately detect and classify the growth and development status of zebrafish eggs; machine vision technology can monitor and analyze the status information such as the viability of zebrafish eggs in real time, making the detection more intelligent and accurate.

[0050] (4) The image acquisition device has excellent measurement performance. The combination of low and high magnification effectively solves the contradiction between the measurement range and measurement accuracy. Low magnification can quickly determine the approximate position and overall situation of zebrafish eggs, while high magnification can obtain their detailed feature information, improving the detection accuracy while ensuring the detection range.

[0051] (5) Non-destructive testing characteristics: The detection process has no impact on the tested zebrafish eggs and does not interfere with their normal growth and development process, meeting the requirements of non-destructive testing of samples in biological experiments and ensuring the authenticity and reliability of experimental results.

[0052] (6) High detection efficiency and accuracy: With the help of AI technology and machine vision technology, functions such as automatic positioning are realized. Compared with traditional manual detection methods, the detection efficiency and accuracy are greatly improved, and the growth and development information of zebrafish eggs can be obtained more quickly and accurately, providing strong support for related research and applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The drawings are included to provide a further understanding of the embodiments and are incorporated into and constitute a part of this specification. The drawings illustrate the embodiments and, together with the description, are used to explain the principles of the present invention. Many of the expected advantages of the embodiments of the present invention and other embodiments will be readily appreciated as they become better understood by reference to the following detailed description. The elements of the drawings are not necessarily to scale with each other. The same reference numerals refer to corresponding like parts.

[0054] Figure 1It is a flowchart of the zebrafish growth and development detection method according to the present application;

[0055] Figure 2 It is a schematic diagram of the front image acquisition sequence of the porous cell culture plate to be tested in the zebrafish growth and development detection method according to the present application;

[0056] Figure 3a It is a diagram of the developmental stages of zebrafish eggs in the normal state in the zebrafish growth and development detection method according to the present application;

[0057] Figure 3b It is a comparison diagram of the developmental stages of zebrafish eggs in the normal state and abnormal state in the zebrafish growth and development detection method according to the present application;

[0058] Figure 3c It is an image processing result diagram of the zebrafish growth and development detection method according to an embodiment of the present application;

[0059] Figure 4 It is a process architecture diagram of the zebrafish growth and development detection method according to an embodiment of the present application;

[0060] Figure 5 It is a structural diagram of the zebrafish growth and development detection system according to a specific embodiment of the present application;

[0061] Figure 6 It is a structural diagram of the image acquisition device in the zebrafish growth and development detection system according to a specific embodiment of the present application;

[0062] Figure 7 It is a structural diagram of the control component and the support component in the image acquisition device of the zebrafish growth and development detection system according to a specific embodiment of the present application;

[0063] Figure 8 It is a structural diagram of the fixing member in the image acquisition device of the zebrafish growth and development detection system according to a specific embodiment of the present application;

[0064] Figure 9 It is a schematic structural diagram of the computer system of the electronic device suitable for implementing the embodiment of the present application.

[0065] Meanings of each number in the figure: 100 - positioning camera, 200 - upright camera assembly, 300 - inverted camera assembly, 400 - control assembly, 500 - multi-well cell culture plate, 600 - support assembly, 700 - stage, 201 - upright camera, 202 - upright light source, 301 - inverted camera, 302 - inverted light source, 401 - first stepping motor, 402 - second stepping motor, 403 - Y-axis track 1, 404 - Y-axis track 2, 405 - X-axis track, 406 - bearing rod, 407 - fixing part, 501 - target hole position, 601 - first support frame, 602 - second support, 603 - base, 604 - inverted "L" shaped support, 701 - transparent area, 2011 - upright high magnification lens, 3011 - inverted high magnification lens, 4051 - first X-axis track, 4052 - second X-axis track, 4071 - card slot. Detailed implementation manners

[0066] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0067] Refer to Figure 1 , Figure 1 shows a flowchart of the zebrafish growth and development detection method of the present application. As shown in the figure, the method includes the following steps:

[0068] S101. Obtain a multi-well cell culture plate to be tested within the field of view of the positioning camera, and calibrate the zero position of the target hole position to be tested on the multi-well cell culture plate to be tested.

[0069] In some specific embodiments, enable the low magnification lens of the positioning camera and its supporting light source, so that the overall contour of the multi-well cell culture plate to be tested and each target hole position to be tested are within the field of view of the low magnification lens. Calibrate the position of the first target hole to be tested on the multi-well cell culture plate to be tested, and set it as the zero position to provide a reference for the precise targeting of the high magnification lens in the subsequent steps.

[0070] S102. Control the upright camera to collect the upright images of the target hole positions to be tested until all the upright images of the target hole positions to be tested in the multi-well cell culture plate to be tested are collected; after re-calibrating the zero position of the target hole position to be tested, control the inverted camera to collect the inverted images of the target hole positions to be tested until all the inverted images of the target hole positions to be tested in the multi-well cell culture plate to be tested are collected. Among them, the upright images and the inverted images contain the image information of the zebrafish tissue.

[0071] In some specific embodiments, the steps of the forward camera or the reverse camera performing autofocus in step S102 include:

[0072] S1021. Continuously collect the original images of the target hole to be measured, and perform Laplacian filtering on the original images;

[0073] S1022. Calculate the gray variance of the image after Laplacian filtering, traverse different focal length positions of the lens in the forward camera or the reverse camera, record the corresponding gray variance change curve, and determine the mechanical position corresponding to the maximum gray variance as the best focus point;

[0074] S1023. Based on the best focus point, complete the acquisition of the forward image or the reverse image of the target hole to be measured.

[0075] Specifically, to achieve autofocus of the forward camera and the reverse camera for the target hole to be measured and obtain the most suitable focal length, the specific process is as follows:

[0076] Autofocus achieves precise focusing by means of real-time analysis of image sharpness. Perform Laplacian filtering on the continuously collected images. This operation uses a second-order differential operator to enhance the edge and texture features in the images. After filtering, calculate the gray variance of the image, which can characterize the dispersion degree of the image pixel values. Since the image sharpness is positively correlated with the gray variance value, that is, the clearer the image, the larger the gray variance value. Traverse different focal length positions of the lens, and record the corresponding gray variance change curve at the same time. Based on this curve, determine the mechanical position corresponding to the maximum gray variance, and this position is the best focus point. Finally, drive the motor for closed-loop positioning to make the lens accurately reach the best focus point, so as to provide clear and accurate imaging conditions for subsequent image acquisition. After the forward camera completes the focusing and image acquisition of all target holes to be measured, recalibrate the zero position of the target hole to be measured, and then use the same autofocus process to control the reverse camera for focusing and image acquisition until the reverse image acquisition of all target holes to be measured in the multi-well cell culture plate to be measured is completed, and both the forward image and the reverse image contain the image information of zebrafish.

[0077] Further, referring to Figure 2 , Figure 2 shows a schematic diagram of the forward image acquisition sequence of the multi-well cell culture plate to be measured according to the zebrafish growth and development detection method of the present application. As shown in the figure, the target holes to be measured in the multi-well cell culture plate to be measured include A1, A2,..., A n and B1, B2,..., B mThe target well positions to be measured in an n-row and m-column format. After calibrating the zero position of the target well position A1B1 in the first row and first column of the multi-well cell culture plate to be measured, switch to the high-magnification lens of the upright camera and control the first stepping motor to move the high-magnification lens at a predetermined step size in the Y-axis direction. When the high-magnification lens completes the acquisition of the upright image of the target well position A1B1, the first stepping motor controls the high-magnification lens to move to the target well position A2B1 in the second row and first column at the predetermined step size in the Y-axis direction, and complete the acquisition of the upright image of the target well position A2B1, until the acquisition of the upright images of all the target well positions in the first column of the multi-well cell culture plate to be measured is completed. Then control the second stepping motor to move the high-magnification lens at a predetermined step size in the X-axis direction. When the high-magnification lens completes the acquisition of the upright image of the target well position A n B2 in the second column of the nth row, the first stepping motor continues to control the high-magnification lens to move to the target well position A n-1 B2 in the second column of the (n - 1)th row and continue to complete the acquisition of the upright image of the second target well position A n-1 B2 in the second column, until the acquisition of the upright images of all the target well positions in the second column of the multi-well cell culture plate to be measured is completed, and traverse all the target well positions in the multi-well cell culture plate to be measured until all the target well positions in the multi-well cell culture plate to be measured are obtained. Before the acquisition of the inverted image, by controlling the first stepping motor and the second stepping motor, the positions of the high-magnification lenses of the upright camera and the inverted camera are reset to zero, and repositioned with the target well position A1B1 in the first row and first column as the reference. Similarly, traverse and acquire the inverted images of all the target well positions one by one, so as to acquire the images of the zebrafish egg states at different angles, ensure the acquisition of comprehensive and multi-perspective zebrafish egg image data, and provide rich materials for subsequent analysis and research.

[0078] S103. Calculate the proportion of zebrafish tissues in the upright image and the inverted image corresponding to each target well position to be measured, and select the upright image or the inverted image with a larger proportion as the image to be measured.

[0079] In some specific embodiments, for the upright image and the inverted image acquired for each target well position to be measured, only one of the images is selected for subsequent detection or training. By calculating the proportion of specific zebrafish tissues in the upright image and the inverted image to the entire zebrafish, the image with the largest proportion value is selected for the subsequent network training and detection process. This is because a large proportion value indicates that the posture of the zebrafish is more stretched under the lens, or the area facing the lens is larger, which can provide more representative and feature-rich image data for model training and detection.

[0080] S104. Use the trained zebrafish growth and development detection model to obtain the growth and development classification information of zebrafish in the target well position to be measured based on the image to be measured.

[0081] In some specific embodiments, the steps for constructing the zebrafish growth and development detection model in step S104 include:

[0082] S1041. Collect images of the zebrafish development process, label the normal state, abnormal state, and corresponding development types in each image, and construct an image dataset of zebrafish growth and development classification information. Among them, the images include top-view images and bottom-view images of zebrafish. The development types in the normal state include zygote stage, cleavage stage, blastula stage, gastrula stage, somite stage, primordium stage, hatching stage, and larva. The development types in the abnormal state include egg coagulation, somite formation failure, and tail non-separation. For details, see Figure 3a and Figure 3b .

[0083] Specifically, collect image data during the zebrafish development process, label the normal state, abnormal state, and different development stages in each image. The normal state includes different egg development stages. The abnormal states include egg coagulation, somite formation failure, and tail non-separation. To improve the generalization ability of the model, use data augmentation techniques (such as rotation, translation, scaling, mirroring) to expand the dataset (for details, see Figure 3c ). It should be particularly noted to ensure that data augmentation does not change the image label information, that is, the augmentation operation cannot damage the position of the classification and detection boxes. Using the standard annotation format of YOLO, each picture requires a.txt file, and each line represents a box, specifically including the class number, the center coordinates of the border (normalized to the [0,1] interval), and the width and height of the border (normalized).

[0084] S1042. Use the YOLOV8 object detection algorithm to construct a binary classification model for detecting zebrafish growth and development, and based on the image dataset, train and optimize the network parameters in the model to obtain a zebrafish growth and development detection model. The zebrafish growth and development detection model includes a detection head, an EfficientNet feature extractor, and a feature fusion module combined by a feature pyramid network and a path aggregation network.

[0085] Specifically, step S1042 includes the following sub-steps:

[0086] Step 10421, construct the YOLOv8 model structure. This model uses EfficientNet to replace the original feature extractor (Backbone) of yolov8. The role of feature extraction is to extract different levels of features from the input image, combines the feature pyramid network (FPN) and the path aggregation network (PAN), realizes the fusion of multi-scale features, and enhances the detection ability of the model on small and large objects. The detection head of YOLOv8 is responsible for generating prediction boxes of different classes and outputting class, bounding box regression, and confidence information.

[0087] Step 10422, Model Training. Environment configuration, install the dependencies of YOLOv8, prepare PyTorch and CUDA (if using GPU acceleration). Model initialization and dataset loading, download the pre-trained model, and select the weights pre-trained on the COCO dataset as the initial model. Use the Dataset class of YOLOv8 to load the training dataset, set the paths of the training and validation sets, and specify the data format. The hyperparameter settings specifically include: the number of training epochs, set to 100 epochs or adjusted appropriately according to the size of the dataset; the batch size, select 8, 16, 32 respectively for training, and select the best result; the size of the input image (imgsz), use an image size of 416x416; the initial learning rate (lr0), set to 0.01; the final coefficient of learning rate decay (lrf), take 0.01.

[0088] Step 10423, Model Optimization. Model optimization includes calculating the loss function, learning rate adjustment, validation and overfitting, hyperparameter optimization, and model evaluation. Among them, the loss function of YOLOv8 includes classification loss, confidence loss, bounding box regression loss, etc., aiming to enable the model to accurately classify objects and precisely regress the object bounding boxes. During training, use a learning rate scheduling strategy (such as Cosine Annealing) to dynamically adjust the learning rate to avoid premature convergence or getting stuck in local optima during training. At the end of each epoch, use the validation set to evaluate the model performance. If the training accuracy improves while the validation accuracy decreases, it indicates that the model may be overfitting, and early stopping, regularization, or reducing the network complexity can be used to solve it. Use cross-validation or grid search to find the best combination of hyperparameters to optimize the classification accuracy and generalization ability.

[0089] Step 10424, Model Evaluation and Accuracy Improvement. Calculate metrics such as accuracy, precision, recall, and F1 value to evaluate the classification performance of the model. For the classification task of the developmental stages of zebrafish, particular attention is paid to the classification accuracy to ensure that each developmental stage is correctly identified. Specifically, when the overall prediction accuracy of the model reaches and stably maintains above 90%, it is determined that the training of the zebrafish growth and development detection model is completed.

[0090] Step 10425, Model Deployment. Export the trained zebrafish growth and development detection model to ONNX or TorchScript format for subsequent deployment or integration into practical applications.

[0091] In some specific embodiments, refer to Figure 4The flowchart of the zebrafish growth and development detection method according to an embodiment of the present application is shown. The method includes:

[0092] Step 401, control software writing and system construction. Use QT to write the motion control software to achieve the control of the hardware motion, and conduct the mechanical structure design and the construction of the optical image acquisition system to achieve the hardware basis for image acquisition.

[0093] Step 402, image acquisition. Based on the above two parts, an automated image acquisition system is jointly formed to perform the image acquisition task, and the zebrafish image to be measured is acquired through this system.

[0094] Step 403, acquire zebrafish images. And perform data augmentation operations on the acquired zebrafish egg images. Specifically, by performing transformations such as translation, rotation, scaling, and mirroring on the original images, the geometric shape and spatial position of the images are changed to achieve the expansion of the data volume. The images after data augmentation and the original images jointly form a complete data set. In addition, for the enhanced images, interpolation algorithms are used for processing to ensure the quality of the images, maintain the integrity and clarity of the image details, and avoid the loss or distortion of image information caused by data augmentation operations. After completing the image acquisition and enhancement processing, operations such as annotation, classification, and abnormal image elimination are performed on the zebrafish egg images. During the annotation process, accurate label information is added to each image according to the characteristics such as the development stage, normal or abnormal state of the zebrafish eggs; classification is to divide the images into corresponding categories according to the established category standards; at the same time, through preset screening rules, abnormal images containing serious noise interference, unclear imaging, or other non-compliant experimental requirements are identified and eliminated to improve the quality and reliability of the data set. The deep learning data set is divided according to the ratio of 75% and 25%. Among them, 75% of the data is used as the training data set for learning and optimizing the network parameters during the model training stage; the remaining 25% of the data is used as the validation data set for evaluating the performance of the model during the training process and monitoring whether the model has overfitting or underfitting phenomena.

[0095] Specifically, collect zebrafish image data at developmental stages including blastula 128-cell stage, somite stage 6 somites, somite stage 18 somites, blastula spherical stage, blastula dome stage, gastrula 30% epiboly, gastrula 75% epiboly, gastrula tail bud stage, hatching stage, primordium stage, etc. Use the annotation tool labelme to label the corresponding developmental stage category for each image. Note that each image can only have one developmental stage label during annotation, as this is a single-label classification problem. The data needs to be annotated in YOLO format, with each image corresponding to a.txt file, which contains: class number (numbered starting from 0), width and height of the image (normalized to the range [0,1]). For example, if there is only one class label "blastula 128-cell stage" in the image, then the label file for this image should only contain the label number of this class.

[0096] Step 404, construct and modify the YOLOV8 model to make it suitable for the zebrafish egg detection task.

[0097] Specifically, the YOLOV8 model uses EfficientNet to replace the original Backbone of yolov8, enhances feature information at different scales through feature fusion (such as FPN / PAN), and the classification head is responsible for generating the classification output of each image to predict the developmental stage to which the image belongs.

[0098] Step 405, YOLOV8 training. Use the YOLOV8 algorithm to train the YOLOV8 model constructed in step 404 using the image dataset constructed in step 403. During the training process, it involves optimizing and adjusting multiple network parameters, including but not limited to batch size (the number of samples input to the model each time during training), β value (a parameter used to control calculations such as momentum in a specific optimization algorithm), optimizer (such as Stochastic Gradient Descent (SGD), Adagrad, Adadelta, Adam, etc., which determines the way the model weights are updated), learning rate (a parameter that controls the step size of the model weight update), etc. Dynamically adjust according to the performance of the model on the training set and validation set.

[0099] Specifically, first, install the dependencies required for YOLOv8, load the pre-trained model (yolov8n.pt), create a data.yaml file, specify the paths of the training set and validation set, as well as the class information. Start training by setting the training parameters. The main training parameters include: the number of training epochs, set to 100 epochs. The batch size, select 8, 16, 32 respectively for training, and select the best result. The size of the input image imgsz, YOLOv8 uses 640x640 as the input size. During the training process, YOLOv8 will automatically select an appropriate learning rate strategy. The initial learning rate (lr0) is usually set to 0.01, and Cosine Annealing is used to dynamically adjust the learning rate. During the training process, YOLOv8 will automatically record metrics such as the loss, accuracy, and recall of each epoch. These metrics can be used to monitor the training progress and save the optimal model at the end of the training.

[0100] Step 406, hyperparameter tuning. To find the best combination of hyperparameters, use automated tools for hyperparameter tuning, such as grid search and Bayesian optimization. Grid search exhaustively traverses all possible combinations within the specified hyperparameter range, evaluates the model performance one by one, and thus determines the optimal parameter combination; Bayesian optimization constructs a probability model of the objective function, uses the existing sample data to predict the value probability of the objective function under different hyperparameter combinations, and guides the search process with this to efficiently find the parameter configuration that makes the objective function reach the optimal value. During the tuning process, hyperparameters such as the learning rate and regularization parameter (a parameter used to prevent model overfitting and constrain the model weights) can also be adjusted to further improve the algorithm performance and enhance the model's detection and classification ability for zebrafish egg images.

[0101] Furthermore, if the model accuracy is not high, cross-validation can be used to further tune the hyperparameters of the model. Adopt regularization methods: L2 regularization, Dropout, etc. to prevent model overfitting.

[0102] Step 407, model deployment. Deploy the tuned model to the microscope software system for actual zebrafish growth and development detection. Export the trained model in ONNX or TorchScript format for convenient subsequent deployment or integration into actual applications.

[0103] For further reference Figure 5 As an implementation of the above method, in a second aspect, the present application provides an embodiment of a zebrafish growth and development detection system 800. This system embodiment is related to Figure 1The method embodiments shown correspond to a system that can be specifically applied to various electronic devices. The system 800 includes a zero-point position calibration module 801, an image acquisition device 802, a to-be-tested image calculation module 803, and a zebrafish development type acquisition module 804 that are communicatively connected to each other, where:

[0104] The zero-point position calibration module 801 is configured to acquire a to-be-tested multi-well cell culture plate within the field of view of a positioning camera and calibrate the zero-point position of a to-be-tested target well position on the to-be-tested multi-well cell culture plate.

[0105] The image acquisition device 802 is configured to control an upright camera to acquire upright images of the to-be-tested target well positions until the acquisition of upright images of all the to-be-tested target well positions within the to-be-tested multi-well cell culture plate is completed;

[0106] After recalibrating the zero-point position of the to-be-tested target well position, control an inverted camera to acquire inverted images of the to-be-tested target well positions until the acquisition of inverted images of all the to-be-tested target well positions within the to-be-tested multi-well cell culture plate is completed, where the upright images and the inverted images contain image information of zebrafish tissues.

[0107] The to-be-tested image calculation module 803 is configured to calculate the occupation ratio of zebrafish tissues in the upright images and the inverted images corresponding to each to-be-tested target well position, and select the upright image or the inverted image with a larger occupation ratio as the to-be-tested image

[0108] The zebrafish development type acquisition module 804 is configured to use a trained zebrafish growth and development detection model to obtain the growth and development classification information of zebrafish in the to-be-tested target well positions based on the to-be-tested images.

[0109] In some specific embodiments, continue to refer to Figure 6 and Figure 7 , Figure 6 and Figure 7The structural diagrams of the image acquisition device, its control component, and the support component of the zebrafish growth and development detection system according to embodiments of the present application are respectively shown. As shown in the figure, the image acquisition device 802 includes a support component 600 and a positioning camera 100, an upright camera assembly 200, an inverted camera assembly 300, a control component 400, and a stage 700 mounted on the support component 600. Among them, a transparent area 701 is provided on the stage 700 to facilitate the transmission of light and meet the requirements of upright and inverted shooting. The positioning camera 100 and the upright camera assembly 200 are located above the transparent area 701, and the inverted camera assembly 300 is arranged below the transparent area 701. The positioning camera 100 uses a low-power lens camera to preliminarily observe and determine the approximate position of the multi-well cell culture plate 500 on the stage 700, providing guidance for the precise positioning of the subsequent high-power lens, facilitating the rapid finding of the target well position to be measured, and narrowing the observation range. The upright camera assembly 200 includes an upright camera 201 and an upright light source 202. The upright camera includes an upright high-power lens 2011. The upright camera 201 is used to capture the top view image of the zebrafish, and the upright light source 202 provides illumination for the shooting to ensure the clarity and contrast of the image. The upright high-power lens 2011 can then perform high-resolution shooting on the target well position to obtain detailed zebrafish egg information. The inverted camera assembly 300 is composed of an inverted camera 301 and an inverted light source 302, and the inverted camera is configured with an inverted high-power lens 3011. The inverted camera 301 is used to collect the bottom view image of the zebrafish, forming a shooting with a different perspective from the upright camera 201 to comprehensively obtain the state information of the zebrafish eggs. The inverted light source 302 provides appropriate lighting conditions for the bottom view shooting, and the inverted high-power lens 3011 is also used for high-resolution shooting. The control component 400 is used to move the upright camera assembly 200 and the inverted camera assembly 300 in a specific direction at a predetermined step length.

[0110] Specifically, the upright camera assembly 200 and the inverted camera assembly 300 are respectively mounted on the first Y-axis track 403 and the second Y-axis track 404 of the control assembly 400 through the fixing member 407. The first Y-axis track 403 and the second Y-axis track 404 are respectively fixed on the same plane perpendicular to the base 603 by the first bracket 601 and the second bracket 602. The control assembly 400 further includes a first stepping motor 401, a second stepping motor 402, and an X-axis track controlled by the second stepping motor 402. The first stepping motor 401 controls the first Y-axis track 403 and the second Y-axis track 404 to move in the Y-axis direction with a predetermined step length, thereby controlling the upright camera assembly 200 and the inverted camera assembly 300 to move in the Y-axis direction with a predetermined step length. The X-axis track includes a first X-axis track 4051 and a second X-axis track 4052 that are parallel to each other and fixed to the base 603. The first X-axis track 4051 and the second X-axis track 4052 are respectively in the same plane as the first bracket 601 and the second bracket 602 of 603. The second stepping motor 402 controls the first X-axis track 4051 and the second X-axis track 4052 to control the plane where the first Y-axis track and the second Y-axis track are located to move in the X-axis direction, thereby controlling the upright camera assembly 200 and the inverted camera assembly 300 to move in the X-axis direction with a predetermined step length. Between the first Y-axis track 403 and the second Y-axis track 404, and between the first X-axis track 4051 and the second X-axis track 4052, connections are respectively achieved through the bearing rods 406. The first stepping motor 402 is connected to the bearing rod between the first Y-axis track 403 and the second Y-axis track 404, and the second stepping motor is connected to the bearing rod between the first X-axis track 4051 and the second X-axis track 4052. When the first stepping motor 402 starts to operate, its power is transmitted through the bearing rod 406, driving the first Y-axis track 403 and the second Y-axis track 404 to perform precise linear motion in the Y-axis direction. Similarly, after the second stepping motor starts, with the transmission of the bearing rod 406, it drives the first X-axis track 4051 and the second X-axis track 4052 to perform precise displacement along the X-axis direction. So that the upright camera assembly 200 and the inverted camera assembly 300 can move smoothly and precisely in the X-Y plane according to the preset program to meet the requirements for camera position adjustment during the zebrafish egg image acquisition process, ensuring the acquisition of high-quality and multi-view zebrafish egg images.

[0111] Specifically, the positioning camera 100 is fixed above the transparent area 701 of the stage 700 through the inverted "L" - shaped bracket 604 of the support assembly 600. One end of the inverted "L" - shaped bracket 604 is fixed on the base 603, and the other end is equipped with a camera with a low - power lens, which is used for preliminary observation and determination of the approximate position of the multi - well cell culture plate 500 on the transparent area 701, providing guidance for the subsequent precise positioning of the high - power lens, facilitating quickly finding the target area and narrowing the observation range.

[0112] Specifically, the image acquisition device is provided with four fixing members 407, which are respectively used to fix the front camera assembly 200, the inverted camera assembly 300, the first bracket 601, and the second bracket 602 to the first Y-axis track 403, the second Y-axis track 404, the first X-axis track 4051, and the second X-axis track 4052. Continuing to refer to Figure 8 , Figure 8 shows a structural diagram of the fixing member of the image acquisition device according to the present application. As Figure 8 shown, the fixing member 407 has a rectangular body structure. Through holes for fixing the front camera assembly 200, the inverted camera assembly 300, the first bracket 601, and the second bracket 602 are provided on its upper surface, and a clamping groove 4071 for fixing to the first Y-axis track 403, the second Y-axis track 404, the first X-axis track 4051, and / or the second X-axis track 4052 is opened on its lower surface.

[0113] In a third aspect, the present application proposes a terminal device, including a processor, a memory, and a computer program stored in the memory. The computer program is executed by the processor to implement the zebrafish growth and development detection method as described in any one of the above.

[0114] In a fourth aspect, the present application proposes a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the zebrafish growth and development detection method as described in any one of the above is implemented.

[0115] Next, referring to Figure 9 , which shows a schematic structural diagram of a computer system 900 suitable for implementing the terminal device or server of the embodiments of the present application. Figure 9 The terminal device or server shown is only an example and should not impose any limitations on the functions and usage scope of the embodiments of the present application.

[0116] As Figure 9 shown, the computer system 900 includes a central processing unit (CPU) 901, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 902 or the program loaded from the storage section 908 into the random access memory (RAM) 903. In the RAM 903, various programs and data required for the operation of the computer system 900 are also stored. The CPU 901, the ROM 902, and the RAM 903 are connected to each other through a bus 904. The input / output (I / O) interface 905 is also connected to the bus 904.

[0117] The following components are connected to the I / O interface 905: an input section 906 including a keyboard, a mouse, etc.; an output section 907 including a liquid crystal display (LCD), etc. and a speaker, etc.; a storage section 908 including a hard disk, etc.; and a communication section 909 including a network interface card such as a LAN card, a modem, etc. The communication section 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to the I / O interface 905 as required. A removable medium 911, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is mounted on the drive 910 as required so that a computer program read therefrom is installed into the storage section 908 as required.

[0118] In particular, according to embodiments of the present disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product that includes a computer program carried on a computer-readable medium, and the computer program includes program code for performing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from a network through the communication section 909, and / or installed from the removable medium 911. When the computer program is executed by the central processing unit (CPU) 901, the above functions defined in the methods of the present application are performed. It should be noted that the computer-readable medium of the present application can be a computer-readable signal medium, a computer-readable medium, or any combination of the two. The computer-readable medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable medium can include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, the computer-readable medium can be any tangible medium that contains or stores a program, and the program can be used by or in combination with an instruction execution system, apparatus, or device. In the present application, the computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, in which the computer-readable program code is carried. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable medium, and the computer-readable medium can send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted by any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0119] Computer program code for performing the operations of this application can be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., by connecting through the Internet using an Internet service provider).

[0120] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in the flowchart or block diagram can represent a module, a program segment, or a part of the code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0121] Although the principles of the present invention have been described in detail above in connection with the preferred embodiments of the present invention, those skilled in the art should understand that the above embodiments are merely illustrative implementations of the present invention and do not limit the scope of the present invention. The details in the embodiments do not constitute a limitation on the scope of the present invention. Without departing from the spirit and scope of the present invention, any obvious changes such as equivalent transformations and simple substitutions based on the technical solutions of the present invention fall within the protection scope of the present invention.

Claims

1. A method for detecting the growth and development of zebrafish, characterized in that: The method comprises: S1, obtaining a multi-porous cell culture plate to be tested within the field of view of a positioning camera, and calibrating the zero position of the target well to be tested of the multi-porous cell culture plate to be tested; S2, controlling the upright camera to collect the upright image of the target well to be measured, until the upright image collection of all the target wells to be measured in the multi-porous cell culture plate to be measured is completed; After recalibrating the zero point position of the target well to be measured, controlling the inverted camera to collect the inverted image of the target well to be measured, until the inverted image collection of all the target wells to be measured in the multi-porous cell culture plate to be measured is completed, wherein the upright image and the inverted image contain image information of the zebrafish tissue; S3, calculating the proportion of zebrafish tissue in the upright image and the inverted image corresponding to each target hole position to be measured, and selecting the upright image or the inverted image with a larger proportion as the image to be measured; S4. Using the trained zebrafish growth and development detection model, based on the image to be tested, obtain the growth and development classification information of the zebrafish in the target hole to be tested.

2. The zebrafish growth and development detection method according to claim 1, characterized in that: The step of automatically focusing the upright camera or the reverse camera in step S2 includes: S21, continuously collecting original images of the target hole positions to be measured, and performing Laplace filtering on the original images; S22, calculating the grayscale variance of the image after Laplace filtering, traversing different focal length positions of the lens in the upright camera or the reverse camera, recording the corresponding grayscale variance change curve, and determining the mechanical position corresponding to the maximum grayscale variance as the best focus point; S23, based on the optimal focus point, completing the acquisition of the upright image or the inverted image of the target hole position to be measured.

3. The zebrafish growth and development detection method according to claim 1, characterized in that: The steps of constructing the zebrafish growth and development detection model in step S4 include: S41, collecting images of the zebrafish development process, marking the normal state, abnormal state and corresponding development type in each image, and constructing an image dataset containing classification information of zebrafish growth and development, wherein the images include top view images and bottom view images of the zebrafish; S42. Use the YOLOV8 target detection algorithm to build a binary classification model for detecting the growth and development of zebrafish, and based on the image data set, train and optimize the network parameters in the model to obtain the zebrafish growth and development detection model.

4. The zebrafish growth and development detection method according to claim 3, characterized in that: Step S41 includes removing abnormal images, enhancing images and normalizing the collected images, wherein the enhanced image processing includes translating, rotating, scaling and mirroring the top view image and / or the bottom view image; Among them, the normal developmental types include zygote stage, cleavage stage, blastocyst stage, gastrula stage, somite stage, primordium stage, hatching stage and fry, and the abnormal developmental types include egg condensation, unformed somite and unseparated tail.

5. The zebrafish growth and development detection method according to claim 3, characterized in that: The step S42 includes: S421, constructing a YOLOv8 model structure, wherein the YOLOv8 model includes a detection head, an EfficientNet feature extractor, and a feature fusion module combined with a feature pyramid network and a path aggregation network, wherein the detection head is used to generate prediction boxes of different categories and output category, bounding box regression, and confidence information; S422, performing model training and optimization, training the hyperparameters of the YOLOv8 model based on the image data set, and optimizing the YOLOv8 model by loss function calculation, learning rate adjustment, and hyperparameter combination optimization, wherein the hyperparameters include the number of training rounds, batch size, input image size, initial learning rate, and learning attenuation coefficient; S423, model evaluation and precision improvement, the classification performance of the model is evaluated by calculating the accuracy, precision and F1 value, until the zebrafish growth and development detection model that meets the preset requirements is obtained.

6. A zebrafish growth and development detection system, characterized in that: The system comprises: A zero point position calibration module is configured to obtain the multi-porous cell culture plate to be tested within the field of view of the positioning camera, and calibrate the zero point position of the target hole to be tested of the multi-porous cell culture plate to be tested; An image acquisition device is configured to control the upright camera to acquire the upright image of the target well position to be measured until the upright image acquisition of all the target well positions to be measured in the multi-well cell culture plate to be measured is completed; After recalibrating the zero point position of the target well to be measured, controlling the inverted camera to collect the inverted image of the target well to be measured, until the inverted image collection of all the target wells to be measured in the multi-porous cell culture plate to be measured is completed, wherein the upright image and the inverted image contain image information of the zebrafish tissue; The image calculation module to be tested is configured to calculate the proportion of zebrafish tissue in the positive image and the inverted image corresponding to each target hole position to be tested, and select the positive image or the inverted image with a larger proportion as the image to be tested; The zebrafish development type acquisition module is configured to use the trained zebrafish growth and development detection model to acquire the growth and development classification information of the zebrafish in the target hole to be tested based on the image to be tested.

7. The zebrafish growth and development detection system according to claim 6, characterized in that: The image acquisition device includes a support component and a positioning camera, a positive camera component, a reverse camera component, a control component and a stage installed on the support component, wherein the positive camera component and the reverse camera component are installed on the upper and lower sides of the stage through the control component, wherein: The stage is provided with a transparent area, and the multi-well cell culture plate to be tested is placed in the transparent area; A positioning camera is used to obtain the position of the multi-well cell culture plate to be tested, so as to achieve zero point position calibration of the target well to be tested; The upright camera assembly includes an upright light source and an upright camera, and is used to capture a top view image of the zebrafish in the target hole to be measured; The inverted camera assembly includes an inverted light source and an inverted camera, and is used to capture an upward-view image of the zebrafish in the target hole to be measured; The control component is used to enable the upright camera component and the reverse camera component to move in a specific direction according to a predetermined step length.

8. The zebrafish growth and development detection system according to claim 7, characterized in that: The control component includes a first stepper motor, a second stepper motor, an X-axis track, and a Y-axis track one and a Y-axis track two located in the same plane. The upright camera component and the inverted camera component are respectively fixed on the Y-axis track one and the Y-axis track two. The first stepper motor is used to realize the movement of the upright camera component and the inverted camera component along the Y-axis direction. The second stepper motor is connected to the X-axis track and is used to control the plane where the Y-axis track one and the Y-axis track two are located to move along the X-axis direction.

9. The zebrafish growth and development detection system according to claim 8, characterized in that: The support assembly includes a first bracket, a second bracket and an inverted "L"-shaped bracket for fixing the positioning camera. The Y-axis track 1 and the Y-axis track 2 are fixed on the X-axis track through the first bracket and the second bracket.

10. A computer-readable storage medium, wherein a computer program is stored in the medium, and when the computer program is executed by a processor, the zebrafish growth and development detection method according to any one of claims 1 to 5 is implemented.

Images