Sclera jaundice analysis method based on images

Abstract

The invention discloses an image-based sclera jaundice analysis method, and relates to the technical field of image recognition, and the method comprises the following steps: S1, carrying out image data collection; s2, segmenting the sclera area; s3, color feature extraction is carried out on the sclera area; s4, performing jaundice degree evaluation through color features; s5, performing health assessment on the jaundice level; and S6, performing long-term health monitoring according to the health assessment. Through combination of a data set of a cloud computing platform and an image feature extraction technology, information of more dimensions is fused into a model training process, and through use of a multi-feature support vector machine model based on image features and historical disease course data, the method not only considers features of a single image, but also combines historical data of a patient, so that the accuracy of a model training model is improved. The model can more comprehensively reflect the change trend of jaundice of a patient, so that the prediction precision of the jaundice degree is effectively improved, and the problem that a traditional SVM method is poor in performance under the complex data condition is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and particularly to an image-based scleral jaundice analysis method. Background Art

[0002] Jaundice is a pathological phenomenon caused by abnormal bilirubin metabolism, usually manifested as yellowing of the sclera, skin and other tissues. The detection method of scleral jaundice has always been an important research direction in the medical field. At present, the scleral jaundice diagnosis method based on visual analysis mostly relies on the subjective experience of doctors. However, due to factors such as individual differences in jaundice degree and changes in image acquisition environment, traditional methods often have certain limitations. Therefore, the image-based scleral jaundice analysis method has become a research hotspot for improving the diagnostic accuracy and automation level; Existing image analysis technologies mainly rely on image feature extraction and machine learning algorithms, such as methods like support vector machine (SVM) and deep learning, for the automated assessment of jaundice degree. By using image processing technology to extract features such as color and texture in the scleral image and combining with machine learning models, quantitative analysis of jaundice can be achieved. However, there are still certain deficiencies in the existing technologies, especially in terms of the accuracy and real-time performance of the jaundice analysis model.

[0003] In the prior art, although support vector machine (SVM) has achieved certain success in the field of image classification, in the prediction of jaundice degree, the classification accuracy and generalization ability of the SVM model still face some challenges. The SVM models in the prior art mostly rely on manually extracted image features and usually can only handle relatively simple linear classification problems. As the complexity of jaundice image features increases, the SVM model often has difficulty providing accurate classification results when facing non-linear features. In addition, SVM is sensitive to the selection of hyperparameters, and different kernel functions, regularization parameters, etc. will affect the performance of the model, resulting in unstable performance under different patients or different imaging conditions. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides an image-based scleral jaundice analysis method to solve the problems raised in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solutions: In the first aspect, an embodiment of the present invention provides an image-based scleral jaundice analysis method, including the following steps: S1. Perform image data acquisition: Use a high-definition camera to acquire image data of the eye to obtain an eye image; S2. Segment the scleral area: Analyze the eye image and segment the scleral area based on the shape, color, and brightness information of the eye; S3. Extract color features of the scleral region: Convert the image of the scleral region into the HSV color space to visually represent the hue, saturation, and brightness of the image, thereby obtaining the color features of the scleral region; S4. Evaluate the degree of jaundice through color features: Based on the color features, set up a support vector machine model to quantitatively evaluate the jaundice degree of the sclera and output the jaundice level; S5. Conduct a health assessment based on the jaundice level: Based on the jaundice level, combine the historical course data of the patient to conduct a health assessment and generate a health assessment result; S6. Conduct long-term health monitoring based on the health assessment: Based on the health assessment result, set up a real-time monitoring system, and use the color features of the patient's scleral image and the health assessment result for long-term tracking and dynamic monitoring to obtain monitoring data.

[0006] To further optimize this technical solution, the support vector machine model in step S4 is specifically a support vector machine model based on color space conversion and multi-feature fusion, and the formula is: ; Where, : The predicted degree of jaundice, representing the classification label of jaundice; : The weight vector of the support vector machine, representing the weighting coefficient of the model for each input feature; : The input feature vector, including the image data features of the patient at the current moment ; : The bias term of the SVM model, controlling the offset of the decision plane; : The sign function, used to convert the linear combination into a discrete class label.

[0007] To further optimize this technical solution, the input feature vector in step S4 represents the features of the current patient's scleral jaundice image, including: color features, texture features, and edge features.

[0008] To further optimize this technical solution, the support vector machine model in step S4 trains the support vector machine based on the training set data to determine the optimal weight vector and bias term , and the training is to maximize the margin between classes, and the following formula is set: .

[0009] To further optimize this technical solution, during the training process of the support vector machine model in step S4, support vectors are the training samples closest to the decision boundary. These samples are crucial for the construction of the classification model, and the decision boundary of the model is represented by the following equation: .

[0010] To further optimize this technical solution, the usage process of the support vector machine model in step S4 is as follows: Feature extraction: Extract multiple color features from the image of the scleral region and the historical medical record data of the patient; Feature mapping and fusion: Map the extracted multiple features to a high-dimensional space through non-linear mapping and perform fusion; Train the support vector machine: Use the high-dimensional features after feature fusion and mapping to train the SVM model to determine the optimal classification hyperplane; Jaundice level prediction: Through the trained support vector machine model, input new image data and relevant historical data of the patient, and the model will output the grading result of jaundice, providing a basis for subsequent jaundice evaluation and monitoring.

[0011] To further optimize this technical solution, the real-time monitoring system in step S6 is set up based on the cloud computing platform, specifically an adaptive adjustment model based on dynamic feature update, where the parameters include: is the input feature vector, including the image data features of the patient at the current moment ; represents the degree of jaundice predicted by the SVM model at time , that is, the calculated in step S4; represents the status update based on real-time monitoring data, reflecting the current monitoring status of the system; is the decay factor in the system used to measure the time interval, reflecting the importance of new input data in the decision-making process; is the global patient dataset obtained based on the cloud computing platform, including historical jaundice data and image data information; represents the adjustment parameter based on real-time monitoring adjustment.

[0012] To further optimize this technical solution, the dynamic monitoring mechanism of the real-time monitoring system in step S6 depends on the real-time updated prediction output and image features , and calculates the adjustment decision through the following formula: ; wherein: is the feature vector based on the current moment , the degree of jaundice and the global patient dataset The calculated adjustment value; is the adjustment parameter of the system at the previous moment.

[0013] To further optimize the technical solution, in the step S6 The calculation formula is: ; : The adjustment value calculated based on the feature vector at the current moment , the degree of jaundice The calculated adjustment value; : The adjustment value calculated based on the global patient dataset including historical jaundice data, image data of other patients, etc.; and : The weight coefficient, indicating the influence weight of the current data and the global data on the final adjustment value.

[0014] To further optimize the technical solution, the use process of the calculated in the step S6 includes the following steps: Obtain real-time data; Calculate the current adjustment parameter; Apply the adjustment parameter ; Feedback and dynamic adjustment; Real-time dynamic response.

[0015] In a second aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and wherein: when the computer program instructions are executed by the processor, the steps of a method for analyzing scleral jaundice based on images as described in the first aspect of the present invention are implemented.

[0016] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, and wherein: when the computer program instructions are executed by the processor, the steps of a method for analyzing scleral jaundice based on images as described in the first aspect of the present invention are implemented.

[0017] Compared with the prior art, the present invention provides a method for analyzing scleral jaundice based on images, having the following beneficial effects: The image-based scleral jaundice analysis method incorporates more dimensional information into the model training process by combining the dataset of the cloud computing platform and image feature extraction technology, thereby improving the classification accuracy. By using a multi-feature support vector machine model based on image features and historical disease course data, this method not only considers the features of a single image but also combines the patient's historical data, enabling the model to more comprehensively reflect the jaundice change trend of the patient. This multi-feature fusion approach helps capture more complex non-linear features, enhancing the robustness and generalization ability of the model, thus effectively improving the prediction accuracy of the jaundice degree and avoiding the problem of poor performance of the traditional SVM method under complex data conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 It is a schematic structural diagram of an image-based scleral jaundice analysis method proposed by the present invention; Figure 2 It is a schematic diagram of the automatic exposure control technical process of an image-based scleral jaundice analysis method proposed by the present invention; Figure 3 It is a schematic diagram of the jaundice level output process of an image-based scleral jaundice analysis method proposed by the present invention; Figure 4 It is a schematic diagram of the adjustment parameter usage process of an image-based scleral jaundice analysis method proposed by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will provide a detailed description of the specific embodiments of the present invention in conjunction with the drawings in the specification.

[0021] Many specific details are set forth in the following description to facilitate a thorough understanding of the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0022] Secondly, the "one embodiment" or "embodiment" referred to herein means a specific feature, structure or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it an embodiment that is separate or selectively mutually exclusive with other embodiments.

[0023] Embodiment 1: Referring to Figures 1 to 4 , this is the first embodiment of the present invention. This embodiment provides an image-based scleral jaundice analysis method, including the following steps: S1. Perform image data acquisition: Use a high-definition camera to acquire image data of the eye to obtain an eye image; During the process of using the high-definition camera for image acquisition, an automatic exposure control technology is used to evaluate the brightness information of the current image in real time and dynamically adjust the exposure parameters according to the brightness change, so as to minimize the situation of overexposure or underexposure of the image and improve the quality of image acquisition; The automatic exposure control model is constructed as follows: represents the pixel brightness value at the coordinate in the image; represents the ideal target brightness value, which is usually preset as a value within a standard brightness range and is set according to the requirements of the scleral region; represents the exposure adjustment coefficient, which is an exposure value automatically adjusted through a feedback mechanism; is the sensitivity coefficient of exposure adjustment, which controls the response speed of exposure adjustment, and its value range is [0,1]; is the weight coefficient of the image brightness change, which is used to adjust the relative influence of the brightness distribution; ; Where: represents the average brightness of all pixels in the image, that is, the average brightness of the current image, is the number of pixels in the image; represents the difference between the average brightness of the current image and the target brightness, which determines whether the exposure needs to be increased or decreased.

[0024] represents the measure of the uniformity of the image brightness distribution. If the image brightness changes greatly (for example, there is a strong light contrast), the value of this item will increase, indicating that more exposure adjustment is needed to adapt to the complex brightness change of the image; Its usage process is as follows: Image brightness evaluation: First, the brightness distribution of the currently captured image is obtained in real time through the camera, and the average brightness of the image is calculated. ; Feedback adjustment: Based on the difference between the average brightness of the current image and the target brightness , the model calculates the exposure adjustment coefficient . If the image brightness is too low, then is a positive value, indicating that the exposure needs to be increased; if the image brightness is too high, then is a negative value, indicating that the exposure needs to be decreased; Brightness change compensation: While adjusting the exposure, the model further compensates for areas with large brightness changes (such as highlight or shadow areas) by introducing the uniformity index of the image brightness distribution to avoid over - adjustment or under - adjustment of the exposure; Exposure adjustment implementation: Finally, automatic exposure control is achieved by adjusting the exposure settings of the camera (such as shutter speed, aperture size, or ISO sensitivity) to ensure that the brightness of the image reaches the target value , thereby ensuring the clarity and color accuracy of the scleral area.

[0025] S2. Segment the scleral area: Analyze the eye image and segment the scleral area based on the shape, color, and brightness information of the eye. The segmented image can be used for subsequent jaundice analysis, removing the interference of other parts of the eyeball to ensure the accuracy of the analysis results; S3. Extract color features of the scleral area: Convert the image of the scleral area to the HSV color space to intuitively represent the hue, saturation, and brightness of the image, thereby obtaining the color features of the scleral area; S4. Evaluate the jaundice degree through color features: Based on the color features, set up a support vector machine model to quantitatively evaluate the jaundice degree of the sclera and output the jaundice level; The support vector machine model in step S4 is specifically a support vector machine model based on color space conversion and multi - feature fusion, and the formula is: ; where, : The predicted jaundice degree, representing the classification label of jaundice; : The weight vector of the support vector machine, representing the weighted coefficient of the model for each input feature; : The input feature vector, containing the image data features of the patient at the current moment ; : The bias term of the SVM model, controlling the offset of the decision plane; : Sign function, used to convert the linear combination into discrete class labels; The input feature vector in step S4 represents the features of the current patient's scleral jaundice image, including: Color feature: The color value of the scleral part in the image; Texture feature: The texture feature extracted through the gray-level co-occurrence matrix, reflecting the detailed information of jaundice; Edge feature: The scleral contour information obtained by the edge detection method (such as Canny edge detection); The support vector machine model in step S4 trains the support vector machine based on the training set data to determine the optimal weight vector and bias term , trained to maximize the margin between classes, set as the following formula: ; During the training process of the support vector machine model in step S4, the support vectors are the training samples closest to the decision boundary. These samples are crucial for the construction of the classification model. The decision boundary of the model is represented by the following equation: ; Through this decision boundary, the SVM can divide the data into different classes.

[0026] For example, when , it is classified as mild jaundice; When , it is classified as severe jaundice; When , that is, on the decision boundary, it is in moderate jaundice. Depending on the changes in the medical scenario, it may indicate that the patient's jaundice condition is developing towards severe or recovering to a mild state.

[0027] The usage process of the support vector machine model in step S4 is as follows: Feature extraction: Extract multiple color features from the image of the scleral area and the patient's historical course data; Feature mapping and fusion: Map the extracted multiple features to a high-dimensional space through non-linear mapping and fuse them; Train the support vector machine: Use the high-dimensional features after feature fusion and mapping to train the SVM model to determine the optimal classification hyperplane; Jaundice level prediction: Through the trained support vector machine model, input new image data and the patient's relevant historical data, and the model will output the grading result of jaundice, providing a basis for subsequent jaundice evaluation and monitoring.

[0028] S5. Conduct a health assessment of the jaundice level: Based on the jaundice level and combined with the patient's historical course data, conduct a health assessment to obtain the health assessment result. By integrating various information, we can provide personalized health advice for each patient, predict the possible impact of jaundice on physical health, and provide corresponding early warning and intervention suggestions according to different jaundice levels; S6. Conduct long-term health monitoring based on the health assessment: Based on the health assessment result, set up a real-time monitoring system, and use the color characteristics of the patient's scleral image and the health assessment result for long-term tracking and dynamic monitoring to obtain monitoring data; The real-time monitoring system in step S6 is set up based on the cloud computing platform, specifically an adaptive adjustment model based on dynamic feature update, where the parameters include: is the input feature vector, including the image data features of the patient at the current moment ; represents the degree of jaundice predicted by the SVM model at time , that is, the calculated in step S4; represents the status update based on the real-time monitoring data, reflecting the current monitoring status of the system; is the decay factor used in the system to measure the time interval, reflecting the importance of new input data in the decision-making process; is the global patient dataset obtained based on the cloud computing platform, including historical jaundice data and image data information; represents the adjustment parameter adjusted based on the real-time monitoring; The dynamic monitoring mechanism of the real-time monitoring system in step S6 depends on the real-time updated prediction output and image features , and calculates the adjustment decision through the following formula: ; where: is the adjustment value calculated based on the feature vector at the current moment, the degree of jaundice and the global patient dataset ; is the decay factor, controlling the importance of new input data; a larger will make the impact of new data on system adjustment greater, and vice versa emphasizes the importance of historical data.

[0029] is the adjustment parameter of the system at the previous moment; In the step S6 The calculation formula is: ; : The adjustment value calculated based on the feature vector at the current moment , jaundice degree ; : The adjustment value calculated based on the global patient dataset including historical jaundice data, image data of other patients, etc.; and : The weight coefficient, indicating the influence weights of the current data and the global data on the final adjustment value; In the step S6, the calculated The usage process includes the following steps: Obtain real-time data: At each moment , the system will obtain image data from the scleral region of the patient , and predict the jaundice degree through the SVM model , that is, steps S3 and S4; Calculate the current adjustment parameter: Calculate through the formula ; Apply the adjustment parameter ; Feedback and dynamic adjustment: As time goes by, the system will continuously update the adjustment parameter according to the image data and jaundice prediction results at each moment . The new adjustment parameter will affect the data acquisition and processing strategy at the next moment. Each new calculation will have a direct impact on the behavior of the system, so as to ensure that the monitoring system can adapt to the changes of jaundice; Real-time dynamic response: In this way, the system can continuously adapt to different situations of the patient's jaundice changes. After each new data acquisition, the system automatically adjusts the image acquisition and processing strategy according to the update of, so as to ensure the accuracy and timeliness of monitoring.

[0030] Among them, applying the adjustment parameter includes: Image acquisition frequency: If the jaundice prediction result indicates severe jaundice, the system may increase the image acquisition frequency to obtain the scleral images of the patient more frequently, so as to ensure timely capture of the changes of jaundice; The monitoring system will be based on to determine the interval time of image acquisition at the next acquisition time , the following relationships are used: ; where is the exposure interval adjustment coefficient; Exposure setting: If the current jaundice prediction result shows mild or severe jaundice, the system may need to increase the exposure to ensure that the quality of the image is clear enough for subsequent analysis; The relationship with the exposure time is as follows: ; is the current exposure time, is the exposure factor (empirical value), is the exposure time adjustment coefficient; Image processing accuracy: The system may need to adjust the image processing accuracy according to the severity of jaundice. For severe jaundice, higher-precision image processing algorithms may be required to improve the sensitivity and accuracy of jaundice detection; The image processing accuracy at the next moment is directly affected by : ; is the precision adjustment coefficient.

[0031] Embodiment 2: This embodiment also provides a computer device, applicable to a situation of an image-based scleral jaundice analysis method, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement an image-based scleral jaundice analysis method as proposed in the above embodiment.

[0032] This embodiment also provides a storage medium, on which a computer program is stored, and when the program is executed by the processor, it implements an image-based scleral jaundice analysis method as proposed in the above embodiment.

[0033] The computer device can be a terminal, which includes a processor, a memory, a communication interface, a display screen, and an input device connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner, and the wireless manner can be achieved through WIFI, carrier network, NFC (Near Field Communication), or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or a button, trackball, or touchpad provided on the outer shell of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0034] If a function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in various embodiments of the present invention. The aforementioned storage medium includes: USB flash drive, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk, or optical disc, etc., various media that can store program codes.

[0035] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch instructions from the instruction execution system, apparatus, or device and execute the instructions), or in combination with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device.

[0036] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection (electronic device) having one or more wirings, a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable media can even be paper or other suitable media on which a program can be printed, as the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing it as appropriate, and then storing it in a computer memory.

[0037] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above-described embodiments, the multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or a combination of the following techniques known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), and the like.

[0038] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. An image-based scleral jaundice analysis method, characterized in that: The following steps are involved: S1. Collecting image data: using a high-definition camera to collect image data of the eye to obtain an eye image; S2. Segment the sclera area: Analyze the eye image and segment the sclera area based on the shape, color and brightness information of the eye; S3, extracting color features of the sclera area: converting the image of the sclera area into the HSV color space, intuitively representing the hue, saturation and brightness of the image, thereby obtaining the color features of the sclera area; S4. Evaluate the degree of jaundice by color features: Based on the color features, set up a support vector machine model to quantitatively evaluate the degree of jaundice of the sclera and output the jaundice grade; S5. Perform health assessment based on jaundice level: Perform health assessment based on jaundice level and historical medical data of the patient, and generate health assessment results; S6. Conduct long-term health monitoring based on health assessment: Based on the health assessment results, a real-time monitoring system is established to use the color characteristics of the patient's sclera image and the health assessment results to conduct long-term tracking and dynamic monitoring to obtain monitoring data.

2. The image-based scleral jaundice analysis method according to claim 1, characterized in that: The support vector machine model in step S4 is specifically a support vector machine model based on color space conversion and multi-feature fusion, wherein the formula is: ； in, : The predicted degree of jaundice, indicating the classification label of jaundice; : The weight vector of the support vector machine, which represents the weight coefficient of the model for each input feature; : Input feature vector, containing the patient's Image data features; : The bias term of the SVM model controls the offset of the decision plane; : symbolic function used to convert linear combinations Convert to discrete category labels.

3. The image-based scleral jaundice analysis method according to claim 2, characterized in that: The feature vector input in step S4 represents the features of the current patient's scleral jaundice image, including: color features, texture features and edge features.

4. The image-based scleral jaundice analysis method according to claim 2, characterized in that: The support vector machine model in step S4 trains the support vector machine based on the training set data to determine the optimal weight vector and bias term. , training is to maximize the interval between categories, setting the following formula: 。 5. The image-based scleral jaundice analysis method according to claim 2, characterized in that: During the training process of the support vector machine model in step S4, the support vector is the training sample closest to the decision boundary, and the decision boundary of the model is represented by the following equation: 。 6. The image-based scleral jaundice analysis method according to claim 2, characterized in that: The use process of the support vector machine model in step S4 is as follows: Feature extraction: Extract multiple color features and the patient's historical medical history data from the image of the sclera area; Feature mapping and fusion: The extracted multiple features are mapped to a high-dimensional space through nonlinear mapping and fused; Training support vector machine: Use high-dimensional features after feature fusion and mapping to train the SVM model and determine the optimal classification hyperplane; Jaundice grade prediction: Through the trained support vector machine model, new image data and relevant historical data of the patient are input, and the model will output the jaundice grade results to provide a basis for subsequent jaundice assessment and monitoring.

7. The image-based scleral jaundice analysis method according to claim 1, characterized in that: The real-time monitoring system in step S6 is set up based on a cloud computing platform, specifically an adaptive adjustment model based on dynamic feature updates, wherein the parameters include: is the input feature vector, which contains the patient's Image data features; Indicates at time The degree of jaundice predicted by the SVM model at the moment, that is, the degree calculated in step S4 ; Indicates status updates based on real-time monitoring data, reflecting the current monitoring status of the system; It is the attenuation factor used to measure the time interval in the system, reflecting the importance of new input data in the decision-making process; It is a global patient data set obtained based on the cloud computing platform, including historical jaundice data and image data information; Indicates the adjustment parameters based on real-time monitoring.

8. The image-based scleral jaundice analysis method according to claim 7, characterized in that: The dynamic monitoring mechanism of the real-time monitoring system in step S6 relies on the prediction output updated in real time. and image features , and the adjustment decision is calculated by the following formula: ； in: is the feature vector based on the current moment , degree of jaundice and the global patient dataset The calculated adjustment value; It is the adjustment parameter of the system at the last moment.

9. The image-based scleral jaundice analysis method according to claim 7, characterized in that: In step S6 The calculation formula is: ； : Based on the feature vector at the current moment , degree of jaundice The calculated adjustment value; : Based on the global patient dataset The calculated adjustment values ​​include historical jaundice data and image data of other patients; and : Weight coefficient, which indicates the influence of current data and global data on the final adjustment value.

10. The image-based scleral jaundice analysis method according to claim 7, characterized in that: The calculated value in step S6 The usage process includes the following steps: Get real-time data; Calculate the current adjustment parameters; Applying tuning parameters ; Feedback and dynamic adjustment; Real-time dynamic response.

Images