CN110751112B - Computer vision-based mouse brain map drawing auxiliary system and method - Google Patents

Abstract

The invention discloses a computer vision-based mouse brain map drawing auxiliary system and a computer vision-based mouse brain map drawing auxiliary method, wherein the system comprises a preprocessing module; the input end of the detection module is connected with the first output end of the pretreatment module, and the neuron cell body is detected; the input end of the registration module is connected with the second output end of the preprocessing module, and registration comparison is carried out; the input end of the identification partitioning module is connected with the output end of the registration module, and the mouse brain microscopic image is partitioned; and the first input end of the mapping module is connected with the output end of the detection module, the second input end of the mapping module is connected with the output end of the identification partition module, and the neuron cell bodies and the mouse brain microscopic image are mapped one by one to complete auxiliary drawing of the mouse brain map. The invention solves the problems of insufficient accuracy and incomplete application of the existing software and deep learning algorithm, and realizes the automatic and semi-automatic operation of the auxiliary system for drawing the brain map of the mouse by means of image decomposition and image feature extraction and detection by means of a computer vision algorithm.

Description

A mouse brain mapping assistance system and method based on computer vision

Technical field

The invention relates to the field of computer vision software, and in particular to a mouse brain atlas drawing assistance system and method based on computer vision.

Background technique

The drawing of brain connection maps is of great significance for in-depth understanding of brain functions, simulating brain neural networks, and developing brain-like artificial intelligence. As the basic unit of neural networks, neurons can be divided into many types based on their morphology, development, connections, functions, and gene expression. Neural circuits formed by the interconnections between different types of neurons are the technical structures that carry various brain functions. In the early days of the development of neuroscience research, because brain maps lacked neuron-level resolution, it was impossible to distinguish neuron types, verify the actual existence of synaptic connections, or distinguish direct from indirect connections. However, with the development of genetic labeling methods and microscopic imaging technology, the specificity and resolution of brain connection mapping using optical imaging have rapidly improved. The rapid development of labeling and imaging technology and the rapid accumulation of data have put forward higher requirements for the software and hardware of data acquisition, storage and analysis, especially the identification and reconstruction of neuron morphology.

Some existing commercial or open source software that assists manual reconstruction (such as Neurolucida, NeuroStudio, Simple Neurite Tracer, neuTube, Virtual Finger, etc.) can provide some basic algorithms for automatic reconstruction, but at the whole-brain scale and cell resolution level The accuracy during reconstruction is insufficient, requiring manual correction of the reconstruction results, and it cannot support terabyte-scale big data processing. There is currently no good software for the automatic identification of neuron cell bodies, so it mainly relies on software-assisted manual operations.

In recent years, deep learning algorithms based on brain-like neural networks have made major breakthroughs in image legends and have begun to be applied to the reconstruction and identification of neurons. However, the current application of these algorithms in neurons is limited to the realization of specific brain areas or specific functions, such as methods specifically for registration of deformed brain images and brain atlases (Patent Publication No. CN103268605A, CN106920228A) or three-dimensional reconstruction of brain atlases. Data set calibration (Patent Publication No. CN108564607A) does not form a complete set of brain mapping auxiliary software.

Contents of the invention

The purpose of the present invention is to provide a mouse brain atlas drawing assistance system and method based on computer vision. This system solves the problems of insufficient accuracy of existing software and incomplete application of deep learning algorithms. Aiming at the special needs of mouse brain mapping, it uses computer vision algorithms to decompose large-scale images and automatically extract and detect image features. Mouse brain mapping aids system visualization, interactivity, and automatic and semi-automatic operation.

In order to achieve the above objectives, the present invention provides a mouse brain atlas drawing assistance system based on computer vision. The system includes a preprocessing module, a detection module, a registration module, an identification partition module and a mapping module. The preprocessing module preprocesses the mouse brain microscopic image; the detection module, the input end is connected to the first output end of the preprocessing module, detects the neuron cell bodies in the mouse brain microscopic image based on the computer vision algorithm; the registration module , the input end is connected to the second output end of the preprocessing module, and the preprocessed mouse brain microscopic image is registered and compared with the standard brain atlas to obtain the characteristics of the mouse brain microscopic image; the input end of the identification partition module is connected to the preprocessing module. The output terminal of the quasi-module is connected to automatically identify the partitions in the standard brain atlas, and partition the mouse brain microscopic image according to the characteristics of the mouse brain microscopic image to obtain the mouse brain microscopic image partitions; the mapping module, the first input terminal is connected to the detection The output end of the module is connected, and the second input end is connected to the output end of the identification partition module. The neuron cell body and the mouse brain microscopic image partition are mapped one by one to obtain the partition information of the mouse brain microscopic image partition, and complete the mouse brain Assisted drawing of maps.

The invention also provides a computer vision-based mouse brain atlas drawing auxiliary method. The method is implemented based on a computer vision-based mouse brain atlas drawing auxiliary system and includes the following steps:

Step 1: Transmit the mouse brain microscopic image to the preprocessing module, perform preprocessing to obtain the preprocessed mouse brain microscopic image, and divide the preprocessed mouse brain microscopic image into two channels for transmission;

Step 2: Transmit the first preprocessed mouse brain microscopic image to the detection module, and detect the neuron cell bodies in the preprocessed mouse brain microscopic image based on the computer vision algorithm;

Step 3: Transmit the second preprocessed mouse brain microscopic image to the registration module, and perform registration comparison with the standard brain atlas to obtain the mouse brain microscopic image characteristics;

Step 4: Transmit the characteristics of the mouse brain microscopic image to the identification partition module. The identification partition module automatically identifies the partitions in the standard brain atlas, and analyzes the mouse brain microscopic image according to the characteristics of the mouse brain microscopic image and the partitions in the standard brain atlas. Partition, obtain the mouse brain microscopic image partition;

Step 5: Transfer the neuron cell bodies and the mouse brain microscopic image partitions to the mapping module to complete the mapping, obtain the partition information of the mouse brain microscopic image partitions, and complete the auxiliary drawing of the mouse brain atlas.

Most preferably, pretreatment also includes the following steps:

Step 1.1: Extract the picture sequence from the mouse brain microscopic image, and calculate the contour Hu moment based on the picture sequence to obtain the contour standard;

Step 1.2: Contour matching and adjustment of the mouse brain microscopic image according to the contour standard to obtain the preprocessed mouse brain microscopic image.

Most preferably, the test also includes the following steps:

Step 2.1: Manually label a limited number of neuron cell bodies in the preprocessed mouse brain microscopic image, and crop the preprocessed mouse brain microscopic image into several mouse brain slice microscopic images with the same pixels;

Step 2.2: Pre-train the multi-layer convolutional neural network through the COCO data set, and fine-tune the pre-trained multi-layer convolutional neural network based on the labeled neuron cell bodies to generate a detection neural network;

Step 2.3: Transmit the microscopic image of the mouse brain slice to the detection neural network, and detect the neuron cell bodies in the microscopic image of the mouse brain slice based on the computer vision algorithm;

Step 2.4: Predict the neuron cell bodies in the preprocessed mouse brain microscopic image based on the neuron cell bodies in the mouse brain slice microscopic image.

Most preferably, detecting neuron cell bodies in microscopic images of rat brain slices also includes the following steps:

Step 2.3.1: Locate the center point of the neuron cell body in the microscopic image of the mouse brain slice and its size as the size standard;

Step 2.3.2: According to the size standard, input the detection neural network based on the target detection algorithm in computer vision for prediction, and obtain the confidence heat map;

Step 2.3.3: Calculate the confidence heat map through two regression algorithm branches, and calculate the neuron cell bodies in the microscopic image of the mouse brain slice.

Most preferably, the detection is an automatic/semi-automatic adjustment, and the automatic/semi-automatic detection of the neuron cell bodies is completed through the detection module/manually respectively.

Most preferably, the registration comparison is to compare the features in the preprocessed mouse brain microscopic image with the features in the standard brain atlas one by one to obtain the difference between the preprocessed mouse brain microscopic image and the standard brain atlas. The features correspond one-to-one to the microscopic image features of the mouse brain.

Most preferably, the one-to-one comparison is to use a grid deformation fine-tuning algorithm to fine-tune the features in the pre-processed mouse brain microscopic image according to the features in the standard brain atlas, so that the pre-processed mouse brain microscopic image has Features correspond to features in standard brain atlases.

Most preferably, the registration comparison is based on the interaction of the size of the mouse brain microscopic image partitions with the accuracy level of the resolution of the standard brain atlas.

Most preferably, the drawn mouse brain atlas is a three-dimensional mouse brain atlas model; the auxiliary drawing of the mouse brain atlas based on the mouse brain microscopic image partition is completed at any angle in the three-dimensional model.

The use of this invention solves the problems of insufficient accuracy of existing software and incomplete application of deep learning algorithms. In response to the special needs of mouse brain atlas drawing, computer vision algorithms are used to decompose large-scale images and automatically extract and extract image features. Detection, realizing the functions of visual, interactive, automatic and semi-automatic operation of the mouse brain atlas drawing assistance system.

Compared with the existing technology, the present invention has the following beneficial effects:

1. The mouse brain atlas drawing assistance system provided by the present invention solves the problems of insufficient accuracy of existing software and incomplete application of deep learning algorithms.

2. The mouse brain atlas drawing auxiliary system provided by the present invention is based on the expert knowledge of cranial nerve researchers. It uses computer vision algorithms to decompose large-scale images and extract and detect automatic image features, realizing mouse brain mapping. The automatic and semi-automatic operation of the map drawing assistance system reduces the redundant and boring manual burden of researchers in brain map drawing and provides effective auxiliary functions.

3. The mouse brain atlas drawing auxiliary system provided by the present invention realizes the visual and interactive functions of the mouse brain atlas drawing auxiliary system.

Description of the drawings

Figure 1 is a schematic structural diagram of each module in the mouse brain atlas drawing assistance system provided by the present invention;

Figure 2 is a schematic flow chart of the mouse brain atlas drawing auxiliary method provided by the present invention;

Figure 3 is a schematic flow chart of preprocessing the mouse brain spectrum provided by the present invention;

Figure 4 is a schematic flow chart of mouse brain microscopic image detection provided by the present invention;

Figure 5 is a schematic flow chart of the microscopic image detection of mouse brain slices provided by the present invention.

Detailed ways

The present invention will be further described below through specific embodiments in conjunction with the accompanying drawings. These embodiments are only used to illustrate the present invention and are not intended to limit the scope of protection of the present invention.

The invention is a mouse brain atlas drawing assistance system based on computer vision. It is a visual and interactive software system implemented through the assistance of machine learning methods and expert knowledge.

As shown in Figure 1, the system includes a preprocessing module, a detection module, a registration module, an identification partition module and a mapping module. The preprocessing module preprocesses the mouse brain microscopic image; the detection module, the input end is connected to the first output end of the preprocessing module, detects the neuron cell bodies in the mouse brain microscopic image based on the computer vision algorithm; the registration module , the input end is connected to the second output end of the preprocessing module, and the preprocessed mouse brain microscopic image is registered and compared with the standard brain atlas to obtain the characteristics of the mouse brain microscopic image; the input end of the identification partition module is connected to the preprocessing module. The output end of the quasi-module is connected to automatically identify the partitions in the standard brain atlas, and partition the mouse brain microscopic image according to the characteristics of the mouse brain microscopic image to obtain the mouse brain microscopic image partitions; the mapping module, the first The input end is connected to the output end of the detection module, and the second input end is connected to the output end of the identification partition module. The neuron cell body and the mouse brain microscopic image partition are mapped one by one to obtain the partition information of the mouse brain microscopic image partition. , to complete the auxiliary drawing of mouse brain atlas.

The system has four basic functions; the four basic functions include registration function, automatic identification partition function, detection function and three-dimensional modeling function; these four basic functions can assist researchers in brain mapping and reduce the size of the whole brain The manual complexity of brain mapping can also provide only map auxiliary information for specific brain areas of concern as needed.

The present invention also proposes a mouse brain atlas drawing auxiliary method based on computer vision. The method is implemented based on a mouse brain atlas drawing auxiliary system based on computer vision. As shown in Figure 2, it includes the following steps:

Step 1: Transmit the mouse brain microscopic image to the preprocessing module for preprocessing, obtain the preprocessed mouse brain microscopic image, and divide the preprocessed mouse brain microscopic image into two channels for transmission.

Differences in the structure and imaging technology of the collected mouse brain microscopic images will cause distortion and deviation of the mouse brain atlas. Preprocessing is required to prepare for the registration calculation with the standard brain atlas in the input registration module. As shown in Figure 3, preprocessing also includes the following steps:

Step 1.1: Extract the picture sequence from the mouse brain microscopic image, and calculate the contour image moment (Hu moment) based on the picture sequence in the mouse brain microscopic image to obtain the contour standard;

Step 1.2: Perform contour matching and adjustment on the mouse brain microscopic image according to the contour standard, and select the preprocessed mouse brain microscopic image.

Step 2: Transmit the first preprocessed mouse brain microscopic image to the detection module, and detect the neuron cell bodies in the preprocessed mouse brain microscopic image based on the computer vision algorithm; as shown in Figure 4, Testing also includes the following steps:

Step 2.1: Manually label a limited number of neuron cell bodies in the preprocessed mouse brain microscopic image, and crop the preprocessed mouse brain microscopic image into several mouse brain slice microscopic images with the same pixels ;

Step 2.2: Pre-train the multi-layer convolutional neural network layer through the Microsoft Common Objects inContext (COCO data set), and perform multi-layer convolutional neural network training based on the neuron cell bodies in the labeled mouse brain microscopic images. The network layer is fine-tuned to generate a detection neural network, so that the detection neural network has a relatively high recognition accuracy for neuron cell bodies in mouse brain microscopic images with a small amount of data sets.

Step 2.3: Transmit the microscopic image of the mouse brain slice to the detection neural network, and detect the neuron cell bodies in the microscopic image of the mouse brain slice based on the computer vision algorithm; detecting the neuron cell bodies in the microscopic image of the mouse brain slice also includes the following steps:

Step 2.3.1: The size of the mouse brain microscopic image is large. The number of neuron cell bodies in the mouse brain microscopic image to be identified for detection is large and the area in the mouse brain microscopic image is small. Crop the mouse brain microscopic image into several mouse brain slice microscopic images with the same pixels to improve the accuracy of neuron cell body detection in the mouse brain slice microscopic image; and take any one of the mouse brain slice microscopic images to locate the neurons in the mouse brain slice microscopic image The center point of the cell body and its size serve as size standards;

Step 2.3.2: According to the size standard, input the detection neural network into the detection neural network for prediction based on the target detection algorithm in computer vision, and obtain the confidence heat map;

Step 2.3.3: Calculate the confidence heat map through two regression algorithm branches, and calculate the neuron cell bodies in the microscopic image of the mouse brain slice.

Step 2.4: Predict the neuron cell bodies in the preprocessed mouse brain microscopic image based on the neuron cell bodies in the mouse brain slice microscopic image.

Predicting the neuronal cell bodies in the preprocessed rat brain microscopic images is calculated based on the size of the neuronal cell bodies in the rat brain slice microscopic images.

The detection is automatically/semi-automatically adjusted, and the automatic/semi-automatic detection of neuron cell bodies in the preprocessed mouse brain microscopic images is completed through the detection module/manually respectively.

Step 3: In order to realize the functional identification and tracking function of neuron cell bodies in the mouse brain microscopic image, it is necessary to first clarify the partition of the mouse brain atlas to which the neuron cell bodies in the mouse brain microscopic image belong, so the second path is preset. The processed mouse brain microscopic image is transmitted to the registration module and compared with the standard brain atlas to obtain the mouse brain microscopic image characteristics.

Step 4: Transmit the characteristics of the mouse brain microscopic image to the identification partition module. The identification partition module automatically identifies the partitions in the standard brain atlas, and analyzes the mouse brain microscopic image according to the characteristics of the mouse brain microscopic image and the partitions in the standard brain atlas. Partitioning, obtaining the partitions of mouse brain microscopic images.

Registration comparison is to compare the features in the preprocessed mouse brain microscopic image with the features in the standard brain atlas one by one, and obtain the features in the preprocessed mouse brain microscopic image and the standard brain atlas one by one. Corresponding microscopic image features of mouse brain.

Among them, the one-to-one comparison uses a grid deformation fine-tuning algorithm to fine-tune the features in the pre-processed mouse brain microscopic image according to the features in the standard brain atlas, so that the features in the pre-processed mouse brain microscopic image are consistent with the features in the standard brain atlas. Corresponds to features in standard brain atlases.

The registration calculation uses a machine learning algorithm, and according to the different complexity requirements of brain atlas drawing, different algorithm models are adopted based on the comprehensive consideration of algorithm accuracy and operating efficiency when facing large-scale mouse brain images.

Registration comparison is a fundamental interaction between the size of the mouse brain microscopic image partitions and the accuracy level of the resolution of the standard brain atlas. The operation on the standard brain atlas is performed at the accuracy of the resolution of the standard brain atlas, and the operation on the mouse brain microscopic image is performed at the resolution of the mouse brain microscopic image. The two are matched through the image registration module. The accuracy of the standard brain atlas and mouse brain microscopic images is not lost in the registration process.

Step 5: Transfer the neuron cell bodies and mouse brain microscopic image partitions in the mouse brain microscopic image to the mapping module to complete the mapping, obtain the partition information of the mouse brain microscopic image partition, and complete the auxiliary drawing of the mouse brain atlas.

The drawn mouse brain atlas is a three-dimensional mouse brain atlas model; the auxiliary drawing of the mouse brain atlas based on the mouse brain microscopic image partition is completed at any angle in the three-dimensional model, which can expand the viewing angle range of the standard brain atlas and simulate the mouse brain atlas. Angle distortion caused by brain microscopic images during detection.

Working principle of the invention:

Transmit the mouse brain microscopic image to the preprocessing module, perform preprocessing to obtain the preprocessed mouse brain microscopic image, and divide the preprocessed mouse brain microscopic image into two channels for transmission; The mouse brain microscopic image is transmitted to the detection module, and the neuron cell body in the preprocessed mouse brain microscopic image is detected based on the computer vision algorithm; the second preprocessed mouse brain microscopic image is transmitted to the registration module , register and compare with the standard brain atlas to obtain the mouse brain microscopic image features; transfer the mouse brain microscopic image features to the recognition partition module, and the recognition partition module automatically identifies the partitions in the standard brain atlas, and based on the mouse brain display The microimage features and the partitions in the standard brain atlas are used to partition the mouse brain microscopic image to obtain the mouse brain microscopic image partitions; the neuron cell body and the mouse brain microscopic image partitions are transmitted to the mapping module to complete the mapping, and the mouse brain microscopic image partitions are obtained. The partition information of the micro-image partition completes the auxiliary drawing of the mouse brain atlas.

In summary, the present invention is a computer vision-based mouse brain atlas drawing auxiliary system and method, which solves the problems of insufficient accuracy of existing software and incomplete application of deep learning algorithms, and addresses the special needs of mouse brain atlas drawing. , with the help of computer vision algorithms for decomposition of large-scale images and automatic extraction and detection of image features, the mouse brain atlas drawing assistance system has the functions of visualization, interactivity, automatic and semi-automatic operation.

Although the content of the present invention has been described in detail through the above preferred embodiments, it should be understood that the above description should not be considered as limiting the present invention. Various modifications and substitutions to the present invention will be apparent to those skilled in the art after reading the above. Therefore, the protection scope of the present invention should be defined by the appended claims.

Claims (8)

1. A computer vision-based mouse brain atlas drawing auxiliary method, characterized in that the method is based on a computer vision-based mouse brain atlas drawing auxiliary system, and the computer vision-based mouse brain atlas drawing auxiliary system is implemented. The map drawing auxiliary system includes: a preprocessing module, which preprocesses the mouse brain microscopic image; a detection module, whose input end is connected to the first output end of the preprocessing module, and which detects the defects in the mouse brain microscopic image based on a computer vision algorithm. The neuron cell body; the registration module, the input end is connected to the second output end of the preprocessing module, and the preprocessed mouse brain microscopic image is registered and compared with the standard brain atlas to obtain the mouse brain microscopic image Features; identifying partition module, the input end is connected to the output end of the registration module, automatically identifies the partitions in the standard brain atlas, and partitions the mouse brain microscopic image according to the characteristics of the mouse brain microscopic image , obtain the mouse brain microscopic image partition; the mapping module, the first input end is connected to the output end of the detection module, the second input end is connected to the output end of the identification partition module, and the neuron cell body is connected to the output end of the detection module. The mouse brain microscopic image partitions are mapped one by one to obtain the partition information of the mouse brain microscopic image partitions, and complete the auxiliary drawing of the mouse brain atlas; The computer vision-based mouse brain mapping auxiliary method includes the following steps: Step 1: Transmit the mouse brain microscopic image to the preprocessing module, perform preprocessing to obtain the preprocessed mouse brain microscopic image, and divide the preprocessed mouse brain microscopic image into two channels for transmission; Step 2: Transmit the first preprocessed mouse brain microscopic image to the detection module, and detect the neuron cell bodies in the preprocessed mouse brain microscopic image based on a computer vision algorithm; Includes the following steps: Step 2.1: Manually label a limited number of neuron cell bodies in the preprocessed mouse brain microscopic image, and crop the preprocessed mouse brain microscopic image into several mouse brain slice displays with the same pixels. microimage; Step 2.2: Pre-train the multi-layer convolutional neural network through the COCO data set, and fine-tune the pre-trained multi-layer convolutional neural network based on the labeled neuron cell bodies to generate a detection neural network; Step 2.3: Transmit the microscopic image of the mouse brain slice to the detection neural network, and detect the neuron cell bodies in the microscopic image of the mouse brain slice based on a computer vision algorithm; detecting the microscopic image of the mouse brain slice The neuron cell body in also includes the following steps: Step 2.3.1: Locate the center point of the neuron cell body in the microscopic image of the mouse brain slice and its size as the size standard; Step 2.3.2: According to the size standard, input the detection neural network based on the target detection algorithm in computer vision for prediction, and obtain a confidence heat map; Step 2.3.3: Calculate the confidence heat map through two regression algorithm branches, and calculate the neuron cell bodies in the mouse brain slice microscopic image Step 2.4: Predict the neuron cell bodies in the preprocessed rat brain microscopic image based on the neuron cell bodies in the rat brain slice microscopic image; Step 3: Transmit the second preprocessed mouse brain microscopic image to the registration module, perform registration comparison with the standard brain atlas, and obtain the mouse brain microscopic image characteristics; Step 4: Transmit the mouse brain microscopic image features to the identification partition module. The identification partition module automatically identifies the partitions in the standard brain atlas, and determines the partitions based on the mouse brain microscopic image features and the standard. The partitions in the brain atlas partition the mouse brain microscopic image to obtain the mouse brain microscopic image partitions; Step 5: Transfer the neuron cell bodies and the mouse brain microscopic image partitions to the mapping module to complete the mapping, obtain the partition information of the mouse brain microscopic image partitions, and complete the auxiliary drawing of the mouse brain atlas. 2. The mouse brain atlas drawing auxiliary method based on computer vision as claimed in claim 1, characterized in that the preprocessing further includes the following steps: Step 1.1: Extract a picture sequence from the mouse brain microscopic image, and calculate the contour Hu moment based on the picture sequence to obtain the contour standard; Step 1.2: Perform contour matching and adjustment on the mouse brain microscopic image according to the contour standard to obtain the preprocessed mouse brain microscopic image. 3. The mouse brain atlas drawing auxiliary method based on computer vision as claimed in claim 2, characterized in that the detection is automatic/semi-automatic adjustment, and the automatic adjustment of the neuron cell body is completed by the detection module/manually respectively. /Semi-automatic detection. 4. The mouse brain atlas drawing auxiliary method based on computer vision as claimed in claim 3, characterized in that the registration comparison is to compare the features in the preprocessed mouse brain microscopic image with the The features in the standard brain atlas are compared one by one to obtain the mouse brain microscopic image features in the preprocessed mouse brain microscopic image that correspond to the features in the standard brain atlas. 5. The mouse brain atlas drawing auxiliary method based on computer vision as claimed in claim 4, characterized in that the one-to-one comparison is based on the characteristics in the standard brain atlas through a grid deformation fine-tuning algorithm. The features in the preprocessed mouse brain microscopic image are fine-tuned so that the features in the preprocessed mouse brain microscopic image correspond to the features in the standard brain atlas. 6. The mouse brain atlas drawing auxiliary method based on computer vision as claimed in claim 5, characterized in that the registration comparison is based on the size of the mouse brain microscopic image partition and the standard brain atlas. Basic interactive operations at a precise level of resolution. 7. The mouse brain atlas drawing auxiliary method based on computer vision as claimed in claim 6, characterized in that the drawn mouse brain atlas is a three-dimensional mouse brain atlas model; it is completed according to the partitioning of the mouse brain microscopic image. The auxiliary drawing of the mouse brain atlas is completed at any angle in the three-dimensional model. 8. A computer vision-based mouse brain atlas drawing auxiliary system for implementing the computer vision-based mouse brain atlas drawing auxiliary method according to any one of claims 1 to 7, characterized in that it includes: Preprocessing module, preprocessing mouse brain microscopic images; A detection module, whose input end is connected to the first output end of the preprocessing module, detects neuron cell bodies in the mouse brain microscopic image based on a computer vision algorithm; A registration module, whose input end is connected to the second output end of the preprocessing module, registers and compares the preprocessed mouse brain microscopic image with the standard brain atlas to obtain the mouse brain microscopic image characteristics; Identification partition module, the input end is connected to the output end of the registration module, automatically identifies the partitions in the standard brain atlas, and partitions the mouse brain microscopic image according to the characteristics of the mouse brain microscopic image, and obtains Mouse brain microscopic image partitions; Mapping module, the first input end is connected to the output end of the detection module, the second input end is connected to the output end of the identification partition module, and the neuron cell body and the mouse brain microscopic image partition are carried out one by one. Mapping, obtaining the partition information of the mouse brain microscopic image partitions, and completing the auxiliary drawing of the mouse brain atlas.