A fully self-service fundus camera control method and control system

Abstract

The present invention provides a fully self-service fundus camera control method and control system, belonging to the technical field of medical devices, and specifically comprising: based on the recognition results of historical fundus images in a group with similar focal lengths, determining the recognition accuracy of suspected disease types in different focal length intervals, and determining the recommended measurement focal length of the user in combination with the comprehensive matching coefficients of different suspected disease types, using the recommended measurement focal length to obtain the fundus image of the user, and when it is determined according to the fundus image that the user has an abnormal disease type whose disease matching result does not meet the requirements, multiple measurement focal lengths are determined based on the recognition accuracy of the abnormal disease type in different focal length intervals, multiple groups of fundus images are obtained using the measurement focal lengths, and when it is determined that the measurement results meet the requirements according to the image clarity and recognition accuracy of the fundus images at different measurement focal lengths, the measurement is stopped, thereby ensuring the analysis and processing efficiency of the fundus images.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical devices, and in particular relates to a fully self-service fundus camera control method and control system. Background Art

[0002] The observation of fundus images can realize auxiliary diagnosis of various diseases such as diabetes, hypertension, coronary heart disease, etc. Therefore, in the existing technical solutions, for example, the invention patent application CN202310955386.0 "Image shooting method based on fundus camera, fundus camera, device and medium" evaluates the user's facial movement and dynamically adjusts the sampling frequency of the fundus image of the fundus camera according to the evaluation results, thereby greatly improving the sampling and analysis efficiency of the fundus image. However, the above technical solutions have the following technical problems:

[0003] When acquiring fundus images, existing technical solutions often use a fixed focal length method to acquire specific fundus images. However, if the fundus image is suspected of having disease manifestation characteristics, if a fixed focal length method is used, it is very likely that the detailed features in the fundus image cannot be accurately acquired, thereby failing to accurately diagnose the user's disease condition.

[0004] Specifically, the present application provides a fully self-service fundus camera control method and control system. Summary of the invention

[0005] To achieve the purpose of the present invention, the present invention adopts the following technical solutions:

[0006] According to one aspect of the present invention, a fully autonomous fundus camera control method is provided.

[0007] A fully self-service fundus camera control method, specifically comprising:

[0008] S1 determines disease matching results of different fundus images according to image features of historical matching fundus images of the user, and determines the suspected disease type and comprehensive matching coefficient of the user using the disease matching results;

[0009] S2 divides the historical fundus images into different focal length similarity groups based on the corresponding focal length intervals of the historical fundus images of the self-service fundus camera, determines the recognition accuracy of the suspected disease types in different focal length intervals based on the recognition results of the historical fundus images in the focal length similarity groups, and determines the recommended measurement focal length of the user in combination with the comprehensive matching coefficients of different suspected disease types;

[0010] S3 uses the recommended measurement focal length to obtain the user's fundus image, and when it is determined based on the fundus image that the user has an abnormal disease type whose disease matching result does not meet the requirements, proceed to the next step;

[0011] S4 determines multiple measurement focal lengths based on the recognition accuracy of the abnormal disease type in different focal length intervals, obtains multiple groups of fundus images using the measurement focal lengths, and stops measuring when the measurement results meet the requirements based on the image clarity and recognition accuracy of the fundus images at different measurement focal lengths.

[0012] The beneficial effects of the present invention are:

[0013] 1. In the present invention, the recommended measurement focal length for the user is determined based on the recognition accuracy of the suspected disease type within different focal length intervals and the comprehensive matching coefficients of different suspected disease types, thereby achieving the determination of the recommended measurement focal length from two perspectives: the recognition accuracy of the suspected disease type and the matching of the suspected disease type of the user, avoiding the technical problem that the measurement accuracy is difficult to meet the requirements due to the use of a unified measurement focal length, and improving the accuracy of recognition processing of eye images.

[0014] 2. In the present invention, whether the measurement result meets the requirements is determined based on the image clarity and recognition accuracy of the fundus image at different measurement focal lengths. Not only the image clarity of the fundus image at the measurement focal length but also the accuracy of the recognition result at the measurement focal length are taken into consideration, thereby achieving an evaluation of the adequacy of the measurement data of the user's fundus image from two perspectives, thereby avoiding the technical problem that the efficiency and accuracy of measurement and recognition cannot meet the requirements due to insufficient data or low clarity.

[0015] A further technical solution is that the historical matching fundus image of the user is determined according to the identity recognition result of the user, and specifically the historical fundus image of the user is used as the historical matching fundus image of the user.

[0016] A further technical solution is that the disease matching result of the fundus image is determined based on the matching between the image features of the fundus image and different disease types, specifically determined by the similarity coefficient between the image features of the fundus image and the image features of different disease types.

[0017] A further technical solution is that the method for determining the suspected disease type of the user is:

[0018] Determining matching coefficients between different historical matching fundus images and different disease types according to the disease matching results, and determining disease matching images of different disease types using the matching coefficients;

[0019] Based on the number of the disease matching images, comprehensive matching coefficients of different disease types are determined, and the suspected disease type is determined using the comprehensive matching coefficients.

[0020] A further technical solution is to divide the historical fundus images into different focal length similarity groups, specifically including:

[0021] The corresponding focal length interval of the historical fundus images is determined according to the corresponding focal lengths of the historical fundus images, and the historical fundus images in the same focal length interval are divided into the same focal length similarity group.

[0022] A further technical solution is that the method for determining the recognition accuracy of the suspected disease type within the focal length interval is:

[0023] Determine the matching coefficient between the image features of different historical fundus images and the suspected disease type using the recognition results of the historical fundus images in the focus-similar group;

[0024] Determine the average value of the matching coefficients of the historical fundus images in the focus-similar group by comparing the matching coefficients with the suspected disease type, and use the average value as the group matching coefficient;

[0025] The group matching coefficient is used as the recognition accuracy rate of the suspected disease type within the focal length interval.

[0026] A further technical solution is that the recognition accuracy of the suspected disease type within the focal length interval ranges from 0 to 1, wherein the higher the recognition accuracy of the suspected disease type within the focal length interval, the higher the accuracy of the historical fundus images within the focal length interval in identifying the suspected disease type.

[0027] A further technical solution is that the method for determining the recognition accuracy of the suspected disease type within the focal length interval is:

[0028] Determine the matching coefficient between the image features of different historical fundus images and the suspected disease type using the recognition results of the historical fundus images in the focus-similar group;

[0029] The accurately identified fundus images in the focus similarity group are determined by the matching coefficient with the suspected disease type, and the recognition accuracy of the suspected disease type within the focal length range is determined by using the proportion of the accurately identified fundus images in the historical fundus images in the focus similarity group.

[0030] A further technical solution is that, when the measurement result does not meet the requirement, the acquisition of the user's fundus image continues until the measurement result meets the requirement.

[0031] In a second aspect, the present application provides a fully self-service fundus camera control system, which adopts the above-mentioned fully self-service fundus camera control method, specifically including:

[0032] Matching coefficient evaluation module, measurement focal length recommendation module, abnormal disease screening module, measurement control module;

[0033] The matching coefficient evaluation module is responsible for determining the disease matching results of different fundus images according to the image features of the user's historical matching fundus images, and determining the suspected disease type and comprehensive matching coefficient of the user using the disease matching results;

[0034] The measurement focal length recommendation module is responsible for dividing the historical fundus images into different focal length similarity groups according to the corresponding focal length intervals of the historical fundus images of the self-service fundus camera, determining the recognition accuracy of the suspected disease types in different focal length intervals based on the recognition results of the historical fundus images in the focal length similarity groups, and determining the recommended measurement focal length of the user in combination with the comprehensive matching coefficients of different suspected disease types;

[0035] The abnormal disease screening module is responsible for acquiring the user's fundus image using the recommended measurement focal length, and determining whether the user has an abnormal disease type whose disease matching result does not meet the requirements based on the fundus image;

[0036] The measurement control module is responsible for determining multiple measurement focal lengths based on the recognition accuracy of the abnormal disease type in different focal length intervals, acquiring multiple groups of fundus images using the measurement focal lengths, and stopping the measurement when the measurement results meet the requirements based on the image clarity and recognition accuracy of the fundus images at different measurement focal lengths.

[0037] Other features and advantages will be described in the following description. The objects and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description and drawings.

[0038] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The above and other features and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings.

[0040] Figure 1 is a flow chart of a fully self-service fundus camera control method;

[0041] Figure 2 is a flow chart of a method for determining a suspected disease type of a user;

[0042] Figure 3 is a flow chart of a method for determining the accuracy of identifying suspected disease types within a focal interval;

[0043] Figure 4 is a flow chart of a method for determining a user's recommended measurement focal length;

[0044] Figure 5 It is a framework diagram of a fully self-service fundus camera control system. DETAILED DESCRIPTION

[0045] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0046] Example 1

[0047] To solve the above problems, according to one aspect of the present invention, Figure 1 As shown, a fully self-service fundus camera control method is provided, which specifically includes:

[0048] S1 determines disease matching results of different fundus images according to image features of historical matching fundus images of the user, and determines the suspected disease type and comprehensive matching coefficient of the user using the disease matching results;

[0049] S2 divides the historical fundus images into different focal length similarity groups based on the corresponding focal length intervals of the historical fundus images of the self-service fundus camera, determines the recognition accuracy of the suspected disease types in different focal length intervals based on the recognition results of the historical fundus images in the focal length similarity groups, and determines the recommended measurement focal length of the user in combination with the comprehensive matching coefficients of different suspected disease types;

[0050] S3 uses the recommended measurement focal length to obtain the user's fundus image, and when it is determined based on the fundus image that the user has an abnormal disease type whose disease matching result does not meet the requirements, proceed to the next step;

[0051] S4 determines multiple measurement focal lengths based on the recognition accuracy of the abnormal disease type in different focal length intervals, obtains multiple groups of fundus images using the measurement focal lengths, and stops measuring when the measurement results meet the requirements based on the image clarity and recognition accuracy of the fundus images at different measurement focal lengths.

[0052] Further, the historical matching fundus image of the user is determined according to the identity recognition result of the user, and specifically, the historical fundus image of the user is used as the historical matching fundus image of the user.

[0053] It is understandable that the disease matching result of the fundus image is determined based on the matching between the image features of the fundus image and different disease types, and is specifically determined by the similarity coefficient between the image features of the fundus image and different disease types.

[0054] Specifically, Figure 2 As shown, the method for determining the suspected disease type of the user is:

[0055] Determining matching coefficients between different historical matching fundus images and different disease types according to the disease matching results, and determining disease matching images of different disease types using the matching coefficients;

[0056] Based on the number of the disease matching images, comprehensive matching coefficients of different disease types are determined, and the suspected disease type is determined using the comprehensive matching coefficients.

[0057] Furthermore, the type of suspected disease is determined by using the comprehensive matching coefficient, specifically including:

[0058] Disease types with comprehensive matching coefficients greater than preset matching coefficients are regarded as suspected disease types.

[0059] It can be understood that when the matching coefficient of the historical matching fundus image is within a preset interval, the historical matching fundus image is determined to be a disease matching image.

[0060] It should also be noted that the method for determining the suspected disease type of the user is:

[0061] Determine matching coefficients between different historical matching fundus images and different disease types according to the disease matching results, and use the matching coefficients to average the matching coefficients between different disease types and historical matching fundus images;

[0062] The average value of the matching coefficients with the historical matching fundus images is used as the comprehensive matching coefficients of different disease types, and the suspected disease type is determined using the comprehensive matching coefficients.

[0063] It should also be noted that the method for determining the suspected disease type of the user is:

[0064] Determine the matching coefficients between different historical matching fundus images and the disease type according to the disease matching result, and when the disease type does not have a historical matching fundus image with a matching coefficient within a preset interval, determine that the disease type does not belong to the suspected disease type;

[0065] When there are historical matching fundus images with matching coefficients within a preset interval for the disease type, the historical matching fundus images with matching coefficients within the preset interval are used as disease matching images. When the proportion of the number of disease matching images is less than the proportion of the number of preset images:

[0066] Acquiring the number of the disease matching images, and when the number of the disease matching images is less than a preset number of images, determining that the disease type does not belong to the suspected disease type;

[0067] When the number of disease matching images is not less than the preset number of images:

[0068] Determine the average value of the matching coefficients of the disease type and the historical matching fundus images by using the matching coefficients, and when the average value of the matching coefficients of the historical matching fundus images meets the requirements, determine that the disease type does not belong to the suspected disease type;

[0069] When the average value of the matching coefficients with the historical matching fundus images does not meet the requirements or when the proportion of the number of disease matching images is not less than the proportion of the preset number of images:

[0070] A comprehensive matching coefficient of the disease type is determined based on the number of the disease matching images and the average value of the matching coefficients with the historical matching fundus images, and the comprehensive matching coefficient is used to determine whether the disease type belongs to a suspected disease type.

[0071] Furthermore, the historical fundus images are divided into different focal length similarity groups, including:

[0072] The corresponding focal length interval of the historical fundus images is determined according to the corresponding focal lengths of the historical fundus images, and the historical fundus images in the same focal length interval are divided into the same focal length similarity group.

[0073] Specifically, it should be noted that Figure 3 As shown, the method for determining the recognition accuracy of the suspected disease type within the focal length interval is:

[0074] Determine the matching coefficient between the image features of different historical fundus images and the suspected disease type using the recognition results of the historical fundus images in the focus-similar group;

[0075] Determine the average value of the matching coefficients of the historical fundus images in the focus-similar group by comparing the matching coefficients with the suspected disease type, and use the average value as the group matching coefficient;

[0076] The group matching coefficient is used as the recognition accuracy rate of the suspected disease type within the focal length interval.

[0077] Furthermore, the recognition accuracy of the suspected disease type within the focal length interval ranges from 0 to 1, wherein the higher the recognition accuracy of the suspected disease type within the focal length interval, the higher the accuracy of the historical fundus images within the focal length interval in identifying the suspected disease type.

[0078] In another embodiment, the method for determining the recognition accuracy of the suspected disease type within the focal length interval is:

[0079] Determine the matching coefficient between the image features of different historical fundus images and the suspected disease type using the recognition results of the historical fundus images in the focus-similar group;

[0080] The accurately identified fundus images in the focus similarity group are determined by the matching coefficient with the suspected disease type, and the recognition accuracy of the suspected disease type within the focal length range is determined by using the proportion of the accurately identified fundus images in the historical fundus images in the focus similarity group.

[0081] Specifically, Figure 4 As shown, the method for determining the user's recommended measurement focal length is:

[0082] Obtaining recognition accuracy rates of suspected disease types in different focal length intervals, and determining corrected recognition accuracy rates of suspected disease types in different focal length intervals based on the product of the recognition accuracy rates and comprehensive matching coefficients of the suspected disease types;

[0083] Determining the average recognition accuracy rate in different focal length intervals according to the average value of the corrected recognition accuracy rates of different suspected disease types in different focal length intervals, and determining the best focal length interval in the focal length intervals using the average recognition accuracy rate;

[0084] The recommended measurement focal length of the user is determined by using the average value of the focal length range corresponding to the optimal focal length interval.

[0085] It can be understood that the optimal focal length interval is the focal length interval with the highest average recognition accuracy.

[0086] In another possible embodiment, the method for determining the user's recommended measurement focal length is:

[0087] Obtaining recognition accuracy rates of suspected disease types in different focal length intervals, and determining corrected recognition accuracy rates of suspected disease types in different focal length intervals based on the product of the recognition accuracy rates and comprehensive matching coefficients of the suspected disease types;

[0088] Determine the suspected disease types whose corrected recognition accuracy rates in different focal length intervals are greater than a preset accuracy threshold, and use them as the accurately identified disease types, and use the number of accurately identified disease types to determine the best focal length interval in the focal length intervals;

[0089] The recommended measurement focal length of the user is determined by using an average value of the corresponding focal lengths of the historical fundus images in the optimal focal length interval.

[0090] In another possible embodiment, the method for determining the user's recommended measurement focal length is:

[0091] Obtain the recognition accuracy of the suspected disease type in different focal length intervals. When the average value of the recognition accuracy of the suspected disease type in the focal length interval does not meet the requirement, it is determined that the focal length interval does not belong to the optimal focal length interval:

[0092] When the average value of the recognition accuracy of the suspected disease type within the focal length interval meets the requirement:

[0093] Determine the number of suspected disease types whose recognition accuracy within the focal length interval is within a preset recognition accuracy interval, and when the number of suspected disease types whose recognition accuracy within the preset recognition accuracy interval does not meet the requirement, determine that the focal length interval does not belong to the optimal focal length interval:

[0094] When the number of suspected disease types within the preset recognition accuracy range meets the requirements:

[0095] The corrected recognition accuracy of the suspected disease type in different focal length intervals is determined based on the product of the recognition accuracy and the comprehensive matching coefficient of the suspected disease type. When the average value of the recognition accuracy of the suspected disease type in the focal length interval does not meet the requirement, it is determined that the focal length interval does not belong to the optimal focal length interval:

[0096] When the average value of the recognition accuracy of the suspected disease type within the focal length interval meets the requirement:

[0097] Determine the suspected disease types whose corrected recognition accuracy rates in different focal length intervals are greater than a preset accuracy threshold, and use them as the accurately identified disease types. When the number of accurately identified disease types in the focal length interval does not meet the requirement, determine that the focal length interval does not belong to the optimal focal length interval:

[0098] When the number of accurately identified disease types within the focal length interval meets the requirement:

[0099] Determining the recognition processing reliability coefficients of different focal length intervals according to the number of accurately recognized disease types in different focal length intervals and the corrected recognition accuracy rates of different suspected disease types, and determining the best focal length interval among the focal length intervals using the recognition processing reliability coefficients;

[0100] The recommended measurement focal length of the user is determined by using an average value of the corresponding focal lengths of the historical fundus images in the optimal focal length interval.

[0101] Furthermore, it is determined that the user has abnormal disease types for which the disease matching results do not meet the requirements, specifically including:

[0102] Determine the matching coefficients between the image features of the fundus image and different suspected disease types, and use the matching coefficients as the fundus image matching coefficients;

[0103] It is determined based on the fundus image matching coefficient whether the user has an abnormal disease type for which the disease matching result does not meet the requirements.

[0104] It should be noted that, when the fundus image matching coefficient of the suspected disease type is greater than a preset matching coefficient threshold, the suspected disease type is determined to be an abnormal disease type whose disease matching result does not meet the requirements.

[0105] Specifically, the method for determining the measured focal length is:

[0106] Sorting different focal length intervals based on the recognition accuracy of the abnormal disease type in different focal length intervals to obtain the order of different focal length intervals;

[0107] Determining the number of focal length intervals to be extracted based on an average value of recognition accuracy rates in different focal length intervals;

[0108] The focal length intervals are screened and determined based on the number and order of the focal length intervals extracted, and the measured focal length is determined based on the average value of the corresponding focal lengths of the historical fundus images within the screened focal length intervals.

[0109] It should be noted that determining whether the measurement results meet the requirements specifically includes:

[0110] Determine the image clarity of the fundus image at different measuring focal lengths according to the image clarity of the fundus image at different measuring focal lengths, and determine the clear fundus image at the measuring focal length using the image clarity;

[0111] Acquire the number of clear fundus images at the measuring focal length, and determine the reliability coefficient of the measurement result at the measuring focal length in combination with the image clarity of different clear fundus images and the recognition accuracy of the measuring focal length;

[0112] The reliability of the measurement result is determined based on the reliability coefficient of the measurement result at different measurement focal lengths, and the reliability of the measurement result is used to determine whether the measurement result meets the requirements.

[0113] Further, when the reliability of the measurement result is within a preset reliability value interval, it is determined that the measurement result meets the requirement.

[0114] It can be understood that when the measurement result does not meet the requirement, the acquisition of the user's fundus image continues until the measurement result meets the requirement.

[0115] Specific examples are given to ensure that the measurement results meet the requirements, including:

[0116] S41 determines the image clarity of the fundus image at different measurement focal lengths based on the image clarity of the fundus image at different measurement focal lengths, and determines whether there is a clear fundus image based on the image clarity of the fundus image, if yes, proceeds to the next step, if no, determines that the measurement result does not meet the requirement;

[0117] S42 uses the image clarity to determine clear fundus images at different measurement focal lengths, and determines whether there is a measurement focal length without a clear fundus image, if so, proceeds to the next step, if not, proceeds to step S45;

[0118] S43: taking the measured focal length without a clear fundus image as an unclear measured focal length, and determining whether the proportion of the unclear measured focal length in the measured focal length meets the requirement, and if so, proceeding to the next step, and if not, determining that the measurement result does not meet the requirement;

[0119] S44 determines weight values ​​of different measurement focal lengths according to recognition accuracy rates of different measurement focal lengths, and determines whether the proportion of the sum of the weight values ​​of the unclear measurement focal lengths in the sum of the weight values ​​of the measurement focal lengths meets the requirement, if so, proceeds to the next step, if not, determines that the measurement result does not meet the requirement;

[0120] S45: obtaining the number of clear fundus images at the measurement focal length, and determining the reliability coefficient of the measurement result at the measurement focal length in combination with the image clarity of different clear fundus images and the recognition accuracy of the measurement focal length, and judging whether the number of measurement focal lengths whose measurement result reliability coefficients do not meet the requirements meets the requirements, if so, proceeding to the next step, if not, determining that the measurement result does not meet the requirements;

[0121] S46 determines the reliability of the measurement result based on the reliability coefficient of the measurement result at different measurement focal lengths, and uses the reliability of the measurement result to determine whether the measurement result meets the requirement.

[0122] Example 2

[0123] Second, as Figure 5 As shown, the present application provides a fully self-service fundus camera control system, which adopts the above-mentioned fully self-service fundus camera control method, specifically including:

[0124] Matching coefficient evaluation module, measurement focal length recommendation module, abnormal disease screening module, measurement control module;

[0125] The matching coefficient evaluation module is responsible for determining the disease matching results of different fundus images according to the image features of the user's historical matching fundus images, and determining the suspected disease type and comprehensive matching coefficient of the user using the disease matching results;

[0126] The measurement focal length recommendation module is responsible for dividing the historical fundus images into different focal length similarity groups according to the corresponding focal length intervals of the historical fundus images of the self-service fundus camera, determining the recognition accuracy of the suspected disease types in different focal length intervals based on the recognition results of the historical fundus images in the focal length similarity groups, and determining the recommended measurement focal length of the user in combination with the comprehensive matching coefficients of different suspected disease types;

[0127] The abnormal disease screening module is responsible for acquiring the user's fundus image using the recommended measurement focal length, and determining whether the user has an abnormal disease type whose disease matching result does not meet the requirements based on the fundus image;

[0128] The measurement control module is responsible for determining multiple measurement focal lengths based on the recognition accuracy of the abnormal disease type in different focal length intervals, acquiring multiple groups of fundus images using the measurement focal lengths, and stopping the measurement when the measurement results meet the requirements based on the image clarity and recognition accuracy of the fundus images at different measurement focal lengths.

[0129] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0130] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0131] The above description is only one or more embodiments of this specification and is not intended to limit this specification. For those skilled in the art, one or more embodiments of this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of this specification shall be included in the scope of the claims of this specification.

Claims

1. A fully self-service fundus camera control method, characterized in that: Specifically include: Determine disease matching results of different fundus images based on image features of historical matching fundus images of the user, and determine the suspected disease type and comprehensive matching coefficient of the user using the disease matching results; The historical fundus images are divided into different focal length similarity groups according to the corresponding focal length intervals of the historical fundus images of the self-service fundus camera, and the recognition accuracy of the suspected disease types in different focal length intervals is determined based on the recognition results of the historical fundus images in the focal length similarity groups, and the recommended measurement focal length of the user is determined in combination with the comprehensive matching coefficients of different suspected disease types; Acquiring the user's fundus image using the recommended measurement focal length, and when determining based on the fundus image that the user has an abnormal disease type for which the disease matching result does not meet the requirements, proceeding to the next step; Based on the recognition accuracy of the abnormal disease type in different focal length intervals, multiple measurement focal lengths are determined, and multiple groups of fundus images are acquired using the measurement focal lengths. When the measurement results meet the requirements based on the image clarity and recognition accuracy of the fundus images at different measurement focal lengths, the measurement is stopped.

2. The fully automated fundus camera control method according to claim 1, wherein: The historical matching fundus image of the user is determined according to the identity recognition result of the user, and specifically, the historical fundus image of the user is used as the historical matching fundus image of the user.

3. The fully automated fundus camera control method according to claim 1, wherein: The disease matching result of the fundus image is determined according to the matching situation between the image features of the fundus image and different disease types, and is specifically determined by the similarity coefficient between the image features of the fundus image and different disease types.

4. The fully automated fundus camera control method according to claim 1, wherein: The method for determining the suspected disease type of the user is: Determining matching coefficients between different historical matching fundus images and different disease types according to the disease matching results, and determining disease matching images of different disease types using the matching coefficients; Based on the number of the disease matching images, comprehensive matching coefficients of different disease types are determined, and the suspected disease type is determined using the comprehensive matching coefficients.

5. The fully automated fundus camera control method according to claim 1, wherein: Divide historical fundus images into different focal length similarity groups, including: The corresponding focal length interval of the historical fundus images is determined according to the corresponding focal lengths of the historical fundus images, and the historical fundus images in the same focal length interval are divided into the same focal length similarity group.

6. The fully automated fundus camera control method according to claim 1, wherein: The method for determining the recognition accuracy of the suspected disease type within the focal length interval is: Determine the matching coefficient between the image features of different historical fundus images and the suspected disease type using the recognition results of the historical fundus images in the focus-similar group; Determine the average value of the matching coefficients of the historical fundus images in the focus-similar group by comparing the matching coefficients with the suspected disease type, and use the average value as the group matching coefficient; The group matching coefficient is used as the recognition accuracy rate of the suspected disease type within the focal length interval.

7. The fully automated fundus camera control method according to claim 6, wherein: The recognition accuracy of the suspected disease type within the focal length interval ranges from 0 to 1, wherein the higher the recognition accuracy of the suspected disease type within the focal length interval, the higher the accuracy of the historical fundus images within the focal length interval in identifying the suspected disease type.

8. The fully automated fundus camera control method according to claim 1, wherein: The method for determining the recognition accuracy of the suspected disease type within the focal length interval is: Determine the matching coefficient between the image features of different historical fundus images and the suspected disease type using the recognition results of the historical fundus images in the focus-similar group; The accurately identified fundus images in the focus similarity group are determined by the matching coefficient with the suspected disease type, and the recognition accuracy of the suspected disease type within the focal length range is determined by using the proportion of the accurately identified fundus images in the historical fundus images in the focus similarity group.

9. The fully automated fundus camera control method according to claim 1, wherein: Confirm that the measurement results meet the requirements, including: Determine the image clarity of the fundus image at different measuring focal lengths according to the image clarity of the fundus image at different measuring focal lengths, and determine the clear fundus image at the measuring focal length using the image clarity; Acquire the number of clear fundus images at the measuring focal length, and determine the reliability coefficient of the measurement result at the measuring focal length in combination with the image clarity of different clear fundus images and the recognition accuracy of the measuring focal length; The reliability of the measurement result is determined based on the reliability coefficient of the measurement result at different measurement focal lengths, and the reliability of the measurement result is used to determine whether the measurement result meets the requirements.

10. A fully autonomous fundus camera control system, using a fully autonomous fundus camera control method according to any one of claims 1 to 9, characterized in that: Specifically include: Matching coefficient evaluation module, measurement focal length recommendation module, abnormal disease screening module, measurement control module; The matching coefficient evaluation module is responsible for determining the disease matching results of different fundus images according to the image features of the user's historical matching fundus images, and determining the suspected disease type and comprehensive matching coefficient of the user using the disease matching results; The measurement focal length recommendation module is responsible for dividing the historical fundus images into different focal length similarity groups according to the corresponding focal length intervals of the historical fundus images of the self-service fundus camera, determining the recognition accuracy of the suspected disease types in different focal length intervals based on the recognition results of the historical fundus images in the focal length similarity groups, and determining the recommended measurement focal length of the user in combination with the comprehensive matching coefficients of different suspected disease types; The abnormal disease screening module is responsible for acquiring the user's fundus image using the recommended measurement focal length, and determining whether the user has an abnormal disease type whose disease matching result does not meet the requirements based on the fundus image; The measurement control module is responsible for determining multiple measurement focal lengths based on the recognition accuracy of the abnormal disease type in different focal length intervals, acquiring multiple groups of fundus images using the measurement focal lengths, and stopping the measurement when the measurement results meet the requirements based on the image clarity and recognition accuracy of the fundus images at different measurement focal lengths.

Images