Animal vital sign monitoring method based on multiple cameras, medium and equipment

Abstract

The invention discloses an animal vital sign monitoring method based on multiple cameras, a medium and equipment, and relates to the technical field of animal vital sign monitoring, the method comprises the following steps: training an animal multi-view video by using a DLC attitude estimation network to obtain key point information; analyzing the mass center motion of the animal according to the key point information of the main view angle, comparing the mass center motion with a motion threshold value, determining the animal state, and obtaining a region of interest in the continuous image sequence of the relative static state time period according to the key point information; detecting angular points in the region of interest, and tracking the angular points to obtain corresponding motion signals; performing correlation analysis on the motion signals, performing sorting based on information gains, and fusing the motion signals higher than a set threshold value to obtain vital sign signals; the vital sign signals are analyzed through a power spectrum, energy distribution of the vital sign signals is obtained, and values of the respiration rate and the heart rate are calculated. According to the invention, manual intervention can be effectively reduced, and animal experiment method innovation and data analysis are promoted.

Description

Technical Field

[0001] The present invention relates to the technical field of animal vital sign monitoring, and in particular, to a method, medium, and device for animal vital sign monitoring based on multiple cameras. Background Art

[0002] Animal experimental research has always been the key to scientific breakthroughs in the fields of medicine and biology. Heart rate and respiratory rate are the most important and basic physiological parameters characterizing the metabolic level of animals, and the obtained data can detect and highly predict dynamic clinical experimental situations. In medicine, it has been found that there is a close relationship between the changes in respiratory rate and some biochemical parameters of blood, which may affect blood oxygen saturation, partial pressure of carbon dioxide, blood oxygen content, etc.

[0003] Previously, in many experimental environments, the respiratory rate was usually measured by manual counting. This method has certain subjectivity and is not a reliable measurement means. With the emergence of automation technology, various remote telemetry technologies have been applied to animal vital sign monitoring, such as invasive devices, non-invasive contact devices, etc. Implanting expensive devices or performing potentially risky physical interventions will cause stress and long-term trauma to animals, seriously affecting the immediate and long-term physiological states of the monitored animals. There may also be an increased risk of spreading infectious diseases between humans and animals or experimental animals during the installation of such monitoring devices. The contact wearable devices may also cause discomfort, skin damage, infection and other risks to the experimental animals. Therefore, non-contact systems for tracking and obtaining animal vital sign data have attracted much attention.

[0004] Non-contact methods such as infrared thermal imaging technology have unclear image details and poor resolution, resulting in relatively low accuracy of the obtained vital sign data. Although the radar-based method has strong penetrability and accurate measurement results, long-term radiation may be harmful to animals and is vulnerable to electromagnetic interference. The method based on imaging photoplethysmography (IPPG) is easily affected by light and cannot be detected at night. For the monitoring method based on computer vision, specific information needs to be extracted from the images for subsequent monitoring and diagnosis. At present, there is still a gap in the information extraction accuracy compared with manual calibration. The animal behavior recognition method based on deep learning requires a large amount of data accumulation and model training to improve the accuracy and reliability of recognition. Summary of the Invention

[0005] The purpose of the present invention is to propose a method for animal vital sign monitoring based on multiple cameras to solve the problem of excessive human intervention in existing animal vital sign monitoring, including the following steps: S1. Select the main view and other views for multi-camera monitoring of animal vital signs, and obtain multi-view videos of the animal; S2. Use the DeepLabCut pose estimation network to train the videos taken from each perspective to obtain the key point information of the animal. S3. Analyze the centroid movement of the animal based on the key point information of the animal in the main perspective and compare it with the movement threshold to determine the animal state. In the continuous image sequence during the relatively static state period, obtain the region of interest according to the key point information. S4. Detect corner points within the region of interest, track the corner points to obtain the corresponding motion signals. S5. Conduct a correlation analysis on the motion signals from multiple perspectives, rank the importance of the relevant components based on information gain, extract the motion signals above the set threshold for fusion to obtain the vital sign signals. S6. Preprocess the vital sign signals, use power spectrum analysis on the preprocessed vital sign signals to obtain the energy distribution of the vital sign signals, obtain the frequency data with continuous energy intensity distribution according to the animal's breathing and heart rate frequency ranges, and calculate the values of the breathing rate and heart rate.

[0006] Furthermore, determining the animal state specifically is as follows: Select a set of key points of the animal in the main perspective, calculate the average value of this set of key points to obtain the centroid position, obtain the animal's motion amplitude in the current frame by comparing the displacement of the centroid in the current frame with that in the previous frame, compare the motion amplitude with the set motion threshold. When the motion amplitude is greater than or equal to the motion threshold, the animal state is motion; when the motion amplitude is less than the motion threshold, the animal state is relatively static.

[0007] Furthermore, the calculation method of the motion threshold is:

[0008]

[0009] Among them, represents the motion threshold, represents the initially set motion threshold, represents the adjustment coefficient, represents the average motion amplitude of the animal within the past n time steps, represents the basic proportionality coefficient, and P represents the body length of the animal.

[0010] Furthermore, the relatively static state period refers to the period of remaining static for more than the length of the vital sign calculation window.

[0011] Furthermore, detect corner points through the Shi-Tomasi algorithm.

[0012] Furthermore, use the KLT algorithm to track the corner points, track the displacement of the corner points between different frames, integrate multi-dimensional information, and obtain multi-channel vital sign signals.

[0013] Further, canonical correlation analysis is used to analyze the correlation of motion signals from multiple perspectives.

[0014] Further, the calculation formula for information gain is:

[0015]

[0016] where represents the information entropy of class C, where C is the class of different respiratory or heart rate states of the animal, represents the value set of the relevant component the value set of, represents the class subset corresponding to when A takes the value v, and are the number of elements in sets C and respectively, represents the class the probability of occurrence.

[0017] For each relevant component , where k is the total number of relevant components, calculate the information gain of each relevant component to form an information gain set . For the information gain set perform a descending order sorting. The sorted information gain order corresponds to the importance sorting of the relevant components. The relevant components with larger information gain are ranked in the front, indicating that they contribute more to distinguishing the animal vital sign classes.

[0018] The present invention also provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the above-mentioned multi-camera-based animal vital sign monitoring method is implemented.

[0019] The present invention also provides an electronic device, including a processor and a memory, the processor is interconnected with the memory, wherein, the memory is used to store a computer program, the computer program includes computer-readable instructions, and the processor is configured to call the computer-readable instructions to execute the above-mentioned multi-camera-based animal vital sign monitoring method.

[0020] The beneficial effects brought by the technical solution provided by the present invention are: The present invention obtains multi-view videos of animals. The method of video stream helps to accelerate data collection and reduce the influence of human intervention in experiments. The DeepLabCut pose estimation network is used to train the videos taken from each view to obtain the key point information of the animals. According to the key point information of the animals in the main view, the centroid movement of the animals is analyzed and compared with the movement threshold to determine the animal state. In the continuous image sequence during the relatively static state period, the region of interest is obtained according to the key point information. Corners are detected within the region of interest, and the corresponding movement signals are obtained by tracking the corners. The correlation analysis of the movement signals from multiple views is performed to obtain the vital sign signals. The vital sign signals are preprocessed, and the power spectrum analysis is used to preprocess the vital sign signals to obtain the energy distribution of the vital sign signals. According to the animal respiration and heart rate frequency ranges, the frequency data with continuous energy intensity distribution is obtained, and the values of the respiration rate and heart rate are calculated. The animal vital signs are recorded from multiple angles and in all directions, and the machine learning method is used to obtain the movement signals and statistically analyze the correlations of various data, effectively reducing human intervention and promoting the innovation of animal experiment methods and data analysis. Description of the Drawings

[0021] Figure 1 is the flowchart of the method for monitoring animal vital signs based on multiple cameras according to an embodiment of the present invention; Figure 2 is the schematic diagram of multi-view shooting of cameras according to an embodiment of the present invention; Figure 3 is the relationship diagram between the centroid displacement of the animal and the threshold according to an embodiment of the present invention; Figure 4 is the ROI selection diagram according to an embodiment of the present invention, where Figure 4 in (a) is the distribution diagram of feature points under the field of view of the main camera, (b) is the right abdominal area under the field of view of the main camera, and (c) is the right abdominal area under the field of view of the second camera; Figure 5 is the block diagram of an electronic device in an exemplary embodiment according to an embodiment of the present invention. Detailed Embodiments

[0022] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be further described below in conjunction with the drawings.

[0023] The flowchart of the method for monitoring animal vital signs based on multiple cameras according to an embodiment of the present invention is as Figure 1 , and specifically includes the following steps: S1. Select the main view and other views for multi-camera monitoring of animal vital signs, and obtain multi-view videos of the animals.

[0024] When selecting the viewing angle for animal monitoring, the first thing to do is to determine the best main monitoring viewing angle, which can comprehensively observe most of the animal's limb characteristics and approximate movement trajectory. From the main viewing angle, the animal's limbs, trunk and other major body parts should be clearly visible, and the animal's movement direction and main activity path within its range of activity can be roughly tracked. For example, for animals that move freely in the experimental field, the main viewing angle should be able to cover all key actions of the animal in its main activity area, whether it is the animal's walking, jumping, or other common behaviors, can be more completely presented from this perspective.

[0025] After determining the main viewing angle, it is also necessary to select other shooting viewing angles at a certain pitch angle. On the one hand, the size of the bird's-eye view should be moderate. If it is too small, it may not be able to obtain enough information and fully observe the overall state of the animal; if it is too large, some details may be lost, or the proportion of the animal in the picture may be too small, affecting the monitoring effect. On the other hand, the selection of other shooting angles should be based on the principle of complementing the main viewing angle. These other viewing angles can observe animals from different directions and heights. For example, shooting from a slightly higher position at a certain angle of pitch can capture the characteristics of the animal's back and the top of the head. At the same time, the animal's vertical movement can also be observed, such as the animal's jumping up and down, lying down and standing up, etc., thereby complementing the information from the main viewing angle to form a more comprehensive and three-dimensional animal activity monitoring picture.

[0026] In actual monitoring, the top view is generally selected as the main view, which is used to determine the animal's movement state and as the basis for selecting ROIs from other viewpoints. Figure 2 shown.

[0027] S2. Use the DeepLabCut posture estimation network to train videos shot from different perspectives to obtain the key point information of animals.

[0028] DeepLabCut is an open source deep learning tool for pose estimation and keypoint tracking. Its core idea is to use convolutional neural networks (CNNs) to identify specific keypoints of objects in images.

[0029] First, collect image and video data containing animals. These data need to cover various postures and angles to ensure the generalization ability of the model. Then, use the graphical user interface of DeepLabCut to manually annotate the images or videos, mainly including the nose, ears, shoulders, limbs, tail, etc. of the animals. Only annotate the key points visible in the current perspective, and do not annotate the occluded or invisible parts. After annotation, use a deep learning model to train the data. The main model used by DeepLabCut is a deep learning model based on a convolutional neural network, usually a pre-trained CNN architecture such as ResNet or VGG. During the training process, the model learns how to accurately identify key points from images. After training, the model needs to be evaluated to ensure its generalization ability on new data. This usually involves using a held-out set or cross-validation to evaluate the performance of the model. After the model is trained and validated, it can be applied to new image and video data to track the selected key points.

[0030] S3. Analyze the centroid movement of the animal based on the key point information of the animal in the main perspective and compare it with the movement threshold to determine the animal state. In the continuous image sequence during the relatively stationary state period, obtain the region of interest according to the key point information.

[0031] Select a set of key points of the animal in the main perspective, calculate the average value of this set of key points to obtain the centroid position, and obtain the movement amplitude of the animal in the current frame by comparing the displacement of the centroid in the current frame with that in the previous frame. Compare the movement amplitude with the set movement threshold. When the movement amplitude is greater than or equal to the movement threshold, the animal state is moving; when the movement amplitude is less than the movement threshold, the animal state is relatively stationary.

[0032] A set of key points of the animal in the main perspective is: , centroid position The calculation formula for

[0033]

[0034] By comparing the centroid of the current frame with the centroid of the previous frame to obtain the animal movement amplitude d, the calculation formula is:

[0035] Compare the movement amplitude d with the set movement threshold T. When , the animal state is moving; when , the animal state is relatively stationary.

[0036] Considering that the differences in the body sizes of different animals will affect their movement ranges, the body length of the animal is used as a reference basis. Assuming the body length of the animal is P, a basic proportionality coefficient (which can be determined according to a large amount of experimental data) is set, and the initially set movement threshold can be expressed as: .

[0037] To more accurately reflect the changes in the animal's movement state, the movement threshold is adjusted by combining the animal's recent movement trend. Let the current time be t , and within the past n time steps, the average movement range of the animal is , which can be obtained by calculating the average value of the displacements of the centroid of the key points of the animal in the main perspective within these n time steps. An adjustment coefficient (determined according to experiments and used to balance the influence degree of the movement trend on the threshold) is introduced, and the calculation formula for the movement threshold is:

[0038] where is the initially set movement threshold. If the recent average movement range of the animal is greater than the basic threshold, it indicates that the animal is in a relatively active state, and it will increase accordingly; conversely, if the recent average movement range is less than the basic threshold, it will decrease, so as to achieve the adaptive adjustment of the movement threshold according to the movement trend.

[0039] In the embodiments of the present invention, 6 key points on the animal's body contour are selected, and the average value of this set of points is calculated to obtain the centroid position. By comparing the displacement of the centroid of the current sequence with that of the previous centroid, the overall movement range of the animal in this sequence is obtained and compared with the preset appropriate movement threshold. When the movement threshold is reached, it can be considered that the animal is moving, and if not, it is stationary. The confirmation of the movement threshold is achieved by calculating the pixel size occupied by the main part of the mouse body, and the threshold for judging the animal's movement is selected at a ratio of 1:0.05. Since the centroid displacement may occasionally have outliers (much higher than the values of nearby points), and the centroid displacement changes continuously in a short period of time, the outliers can be removed from the displacement values. Finally, the continuous image sequence that has remained stationary for more than the length of the vital sign calculation window (corresponding to 10 s) is used as an effective segment for vital sign monitoring. The relationship diagram between the centroid displacement of the animal and the threshold in the embodiments of the present invention refers to Figure 3 .

[0040] During each period of rest, the average distances of the left abdomen and the right abdomen from other perspectives are calculated according to the main perspective. The group with the smaller average distance is selected as the ROI selection side, referring to Figure 4 , Figure 4It is the ROI selection diagram of the embodiment of the present invention, where Figure 4 in (a) is the distribution diagram of feature points under the field of view of the main camera, (b) is the right abdominal area under the field of view of the main camera, and (c) is the right abdominal area under the field of view of the second camera. Figure 4 For the second camera in (a), the right abdomen of the animal should be selected as the ROI selection side, and then the ROI area is selected according to the key point information, such as Figure 4 in (b) and (c).

[0041] S4. Detect corner points within the region of interest and track the corner points to obtain corresponding motion signals.

[0042] Corner points refer to points with drastic gray-level changes in an image, usually corresponding to feature points of an object, such as the edges and textures of an object. Due to their uniqueness and stability in the image, corner points are widely used in fields such as image matching and motion detection. Compared with manual recognition and marking, skin surface feature corner points can be efficiently recognized through automated algorithms, reducing external intervention compared with the manual specific marking of traditional methods and improving the efficiency and accuracy of data processing. Therefore, corner points are suitable for identifying the natural features of the left and right abdomen of animals and serving as signal anchors for vital sign signal acquisition.

[0043] Detect corner points in the selected ROI through the Shi-Tomasi algorithm, and select the strongest 30 points as feature points according to the intensity of the corner points. The Shi-Tomasi algorithm is a corner detection method improved based on the Harris algorithm and is also known as the minimum eigenvalue algorithm. It has good robustness and corner quality evaluation indices, and these advantages enable it to effectively identify the corner point features at various positions of animals and have significant advantages in detecting the feature of the significant region of animal cardiopulmonary function.

[0044] The KLT algorithm is an improved version based on the Lucas-Kanade optical flow algorithm and can accurately track the displacement of the detected key points between different frames. By detailed recording and analysis of the inter-frame displacement of these key points, multi-dimensional information is integrated to obtain initial multi-channel vital sign signals, and these multi-channel signals will reflect the physiological characteristics related to animal cardiopulmonary function in a static state from multiple angles.

[0045] S5. Perform correlation analysis on the motion signals from multiple perspectives, rank the importance of relevant components based on information gain, and extract the motion signals higher than the set threshold for fusion to obtain vital sign signals.

[0046] Using the CCA (Canonical Correlation Analysis) method, the multi-channel signals obtained from the main perspective and other different perspectives are deeply analyzed. The CCA method can effectively mine the internal correlation relationship between two sets of variables. By constructing a suitable mathematical model, the highly correlated components among the multi-channel signals from multiple perspectives are found. Based on these highly correlated components, data fusion is performed to obtain the fused vital sign signals. The fused signals synthesize the key information about the animal's vital signs contained in multiple perspectives and can more comprehensively and accurately reflect the true physiological state of the animal.

[0047] Among them, the calculation formula for information gain is:

[0048]

[0049] Among them, represents the information entropy of category C, where C is the different respiratory or heart rate state categories of the animal, represents the value set of the relevant component , represents the category subset corresponding to when A takes the value v, and are the number of elements in set C and respectively, represents category appearing probability.

[0050] S6. Preprocess the vital sign signals, perform power spectrum analysis on the preprocessed vital sign signals to obtain the energy distribution of the vital sign signals, obtain the frequency data with continuous energy intensity distribution according to the animal's respiratory and heart rate frequency ranges, and calculate the values of the respiratory rate and heart rate.

[0051] The fused vital sign signals are processed by detrending and band-pass filtering to remove most of the noise interference. The power spectrum analysis method is used to deeply analyze the vital sign signals. Power spectrum analysis is a powerful tool for converting time-domain signals into frequency-domain representations, which can reveal the energy distribution of signals at different frequencies. For vital sign signals, physiological activities such as breathing and heartbeat have specific frequency ranges, and through power spectrum analysis, the contributions of these frequency components to the overall signal energy can be clearly seen. The fast Fourier transform (FFT) algorithm is used to calculate the power spectrum of the vital sign signals. The FFT algorithm can efficiently decompose the time-domain signal into a combination of sine and cosine waves at different frequencies, thereby obtaining the amplitude and phase information of the signal at each frequency. By calculating the power spectrum, a frequency spectrum diagram with frequency on the horizontal axis and power (energy) on the vertical axis can be obtained. In this frequency spectrum diagram, the peaks and amplitudes at different frequency positions reflect the energy intensities of the corresponding frequency components in the vital sign signals.

[0052] According to the breathing and heartbeat frequency ranges of animals, frequency data with continuous energy intensity distribution is obtained from the power spectrum. Different species of experimental animals have different breathing and heartbeat frequency ranges. For example, the breathing frequency of adult rats is usually between 80 - 160 breaths per minute (bpm), and the heartbeat frequency is between 250 - 600 bpm. Based on these known frequency ranges, the corresponding frequency intervals are carefully selected in the power spectrum. Within this interval, observe the frequency components with relatively high and continuous energy intensity. These frequency components are often closely related to the breathing and heartbeat activities of animals because breathing and heartbeat are periodic and regular physiological activities in animal vital signs, and they will appear as relatively stable and continuous energy peaks in the power spectrum. The algorithm is used to extract and label these frequency data to ensure that the signal characteristics related to breathing and heartbeat can be accurately captured.

[0053] Finally, the values of the breathing rate and heart rate are calculated based on the obtained frequency data with continuous energy intensity distribution. For the calculation of the breathing rate, within the selected breathing frequency interval, find the frequency value corresponding to the energy peak, and this frequency value is the dominant frequency of the breathing signal. According to the relationship between the breathing frequency and the breathing rate (the breathing rate is equal to the breathing frequency multiplied by 60 seconds), the breathing rate of the animal can be calculated. For example, if the dominant frequency within the breathing frequency interval in the power spectrum is 1.5 Hz (i.e., 90 breaths per minute), then the breathing rate of the animal is 90 breaths per minute. For the calculation of the heart rate, similarly find the frequency corresponding to the energy peak within the heartbeat frequency interval, and then calculate the heart rate value according to the conversion relationship between the heart rate and the heartbeat frequency. During the calculation process, the stability and reliability of the signal are also considered. By averaging the frequency data over multiple consecutive time periods, the influence of accidental factors on the calculation results is reduced, thereby obtaining more accurate breathing rate and heart rate values.

[0054] In an exemplary embodiment, it includes a computer-readable storage medium that stores a computer program. When the computer program is executed by a processor, it implements the above-mentioned multi-camera-based animal vital sign monitoring method.

[0055] Please refer to Figure 5 , in an exemplary embodiment, it further includes an electronic device, which includes at least one processor, at least one memory, and at least one communication bus.

[0056] Among them, a computer program is stored on the memory. The computer program includes computer-readable instructions. The processor calls the computer-readable instructions stored in the memory through the communication bus and executes the above-mentioned multi-camera-based animal vital sign monitoring method.

[0057] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for monitoring animal vital signs based on multiple cameras, characterized in that: The following steps are involved: S1, selecting the main view and other viewpoints of multi-camera monitoring of animal vital signs to obtain multi-view videos of animals; S2. Use the DeepLabCut pose estimation network to train videos shot from different perspectives to obtain the key point information of the animal; S3, analyzing the animal's center of mass movement based on the animal's key point information from the main perspective and comparing it with the motion threshold to determine the animal's state, and obtaining a region of interest based on the key point information in a continuous image sequence in a relatively static state period; S4, detecting corner points in the region of interest, tracking the corner points to obtain corresponding motion signals; S5. Perform correlation analysis on motion signals from multiple perspectives, sort the importance of related components based on information gain, extract motion signals above a set threshold, and fuse them to obtain vital sign signals; S6. Preprocess the vital sign signal, use power spectrum analysis to analyze the preprocessed vital sign signal, obtain the energy distribution of the vital sign signal, obtain continuous frequency data of energy intensity distribution according to the animal's breathing and heart rate frequency range, and calculate the values ​​of breathing rate and heart rate.

2. The method for monitoring animal vital signs based on multiple cameras according to claim 1, characterized in that: Determine the animal's status specifically as follows: Select a group of animal key points from the main perspective, calculate the average value of this group of key points to obtain the center of mass position, obtain the animal motion amplitude of the current frame by comparing the displacement of the center of mass of the current frame with the center of mass of the previous frame, and compare the motion amplitude with the set motion threshold. When the motion amplitude is greater than or equal to the motion threshold, the animal state is moving, and when the motion amplitude is less than the motion threshold, the animal state is relatively still.

3. The method for monitoring animal vital signs based on multiple cameras according to claim 1, characterized in that: The motion threshold is calculated as: T1=k1×P Wherein, T2 represents the motion threshold, T1 represents the initially set motion threshold, and k2 represents the adjustment coefficient. represents the average movement amplitude of the animal in the past n time steps, k1 represents the basic proportional coefficient, and P represents the body length of the animal.

4. The method for monitoring animal vital signs based on multiple cameras according to claim 1, characterized in that: The relative static state period refers to the period of time during which the static state exceeds the length of the vital signs calculation window.

5. The method for monitoring animal vital signs based on multiple cameras according to claim 1, characterized in that: Corner points are detected using the Shi-Tomasi algorithm.

6. The method for monitoring animal vital signs based on multiple cameras according to claim 1, characterized in that: Use the KLT algorithm to track corner points, track the displacement of corner points between different frames, integrate multi-dimensional information, and obtain multi-channel vital sign signals.

7. The method for monitoring animal vital signs based on multiple cameras according to claim 1, characterized in that: Canonical correlation analysis was used to perform correlation analysis on motion signals from multiple viewpoints.

8. The method for monitoring animal vital signs based on multiple cameras according to claim 1, characterized in that: The calculation formula of information gain is: Among them, H(C) represents the information entropy of category C, C is the different breathing or heart rate state categories of animals, Values(A j ) represents the related component A j The value set of C v represents the corresponding category subset when A takes the value v, |C| and |C v | are sets C and C v The number of elements, p(c i ) indicates category c i Probability of occurrence.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

10. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the processor and the memory are interconnected, wherein the memory is used to store a computer program, the computer program comprises computer-readable instructions, and the processor is configured to call the computer-readable instructions to execute the method according to any one of claims 1 to 8.

Images