Hand hygiene management method and system based on image data and storage medium

Abstract

The invention provides a hand hygiene management method and system based on image data, and belongs to the technical field of equipment management, and the method specifically comprises the steps: determining whether a verification image recognition model needs to be constructed or not according to the video stream data of a sensitivity abnormal combination in a monitoring image of a cleaning device of each operating room; according to the method, whether the verification image recognition model needs to perform verification processing on the monitoring image of the similar video stream combination in the cleaning device of the operating room or not is determined according to the video stream data of the sensitivity abnormal combination in the monitoring image of the cleaning device of the operating room, and the reliability degree of monitoring processing of hand hygiene is improved.

Description

Technical Field

[0001] The present invention belongs to the field of image recognition technology, and in particular relates to a hand hygiene management method, system and storage medium based on image data. Background Art

[0002] To monitor and process hand hygiene, CN202510121906.7, "A method for identifying qualified hand hygiene subjects based on video streams," analyzes the video stream, combines action classification sequences and related time statistics to generate a comprehensive score for the hand hygiene process, and associates the score with user identity information to screen out qualified hand hygiene subjects. This enables high-precision, real-time hand hygiene behavior detection and user identity identification. However, all of the above technical solutions have the following technical problems: When monitoring, analyzing and processing hand hygiene, different people have different hand cleaning postures and steps. Therefore, using the same image recognition model may make it difficult for the hand hygiene recognition accuracy to meet the requirements. Therefore, it becomes a technical problem that needs to be solved urgently to construct multiple image recognition models based on the deviations in the hand cleaning postures and steps of different people, and then improve the accuracy of hand hygiene monitoring, analysis and processing.

[0003] To solve the above technical problems, the present application provides a hand hygiene management method, system and storage medium based on image data. Summary of the Invention

[0004] To achieve the purpose of the present invention, the present invention adopts the following technical solutions: Specifically, the present application provides a hand hygiene management method based on image data, which specifically includes: S1 determines similarities among video frames of a hand hygiene video stream based on the analysis results of the hand hygiene monitoring data, divides the video stream into similar video stream combinations based on the similarities, and proceeds to the next step when it is determined that the target type of operating room meets the requirements based on the video stream data of each similar video stream combination in the monitoring image of the operating room cleaning device; S2 determines the abnormal alarm data within each cleaning time interval based on the video stream data in the similar video stream combination, determines the sensitivity abnormal combination in the similar video stream combination based on the abnormal alarm data, and uses the video stream data of the sensitivity abnormal combination in the monitoring image of the cleaning device in each operating room to determine whether it is necessary to construct a verification image recognition model. The video stream data of the sensitivity abnormal combination in the monitoring image of the cleaning device in the operating room is used to determine whether the verification image recognition model needs to perform verification processing on the monitoring image of the similar video stream combination in the cleaning device of the operating room.

[0005] The beneficial effects of the present invention are: Based on the abnormal alarm data, the sensitivity abnormal combination in the similar video stream combination is determined, thereby realizing the screening of the sensitivity abnormal combination from the perspective of the higher alarm rate of cleaning problems in the longer cleaning time interval and the shorter cleaning time interval in the similar video stream combination. It also lays the foundation for determining the reliability of identifying the cleaning status of hand hygiene in each operating room based on the historical video stream data in the monitoring images of the cleaning devices of each operating room according to the sensitivity abnormal combination, and also lays the foundation for whether it is necessary to construct a verification image recognition model.

[0006] The video stream data of the sensitivity abnormal combination in the monitoring image of the cleaning device in the operating room is used to determine whether the verification image recognition model needs to perform verification processing of the monitoring image of a similar video stream combination in the cleaning device in the operating room, thereby avoiding the technical problem of poor reliability of the server's response processing when performing verification processing on the video stream with sensitivity abnormal combination due to the verification processing of the monitoring image of all similar video stream combinations in all operating rooms. By performing verification processing on the monitoring image of similar video stream combinations only in operating rooms where the proportion of video streams with sensitivity abnormal combinations in historical monitoring images is small and the proportion of similar video stream combinations is large, the reliability of the verification processing of similar video stream combinations is guaranteed while also avoiding excessive waste of recognition resources.

[0007] Furthermore, the similarity of the video frames is determined based on the image similarity of different video frames in the video stream, specifically based on the similarity of texture features of the video frames, specifically by calculating the texture feature similarity using a Euclidean distance function.

[0008] Furthermore, the video stream is a single video stream that records the monitoring video during each cleaning process.

[0009] Furthermore, the video stream is divided into similar video stream combinations, specifically including: Based on similarities between the video stream and other video streams in different video frames, determining video frames of the video stream having similar video frames in other video streams as similar video frames; taking a video frame similar to the similar video frame in the other video stream as a corresponding video frame; According to similar video frame data of the video stream and corresponding video frame data in other video streams, it is determined whether the video stream and other video streams can be classified into the same similar video stream combination.

[0010] Further, determining whether verification processing of the monitoring image of the similar video stream combination is required specifically includes: The operating room whose monitoring image of the cleaning device has the similar video stream combination in history is used as the monitoring target operating room, and the number of video streams of the different monitoring target operating rooms in the similar video stream combination is determined based on the historical monitoring data of the monitoring images of the different monitoring target operating rooms; Using historical monitoring data of monitoring images of different monitoring target operating rooms, determining the number of video streams with abnormal sensitivity combinations in different monitoring target operating rooms; Based on the number of video streams with abnormal sensitivity combinations and the number of video streams with similar video stream combinations in different monitoring target operating rooms, the monitoring images of the video streams of the similar video stream combinations in the cleaning device of the operating room are determined, and whether verification processing of the monitoring images of the similar video stream combinations is required.

[0011] In a second aspect, the present invention provides a computer system comprising: a memory and a processor in communication connection, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the above-mentioned hand hygiene management method based on image data when running the computer program.

[0012] In a third aspect, the present invention provides a computer storage medium having a computer program stored thereon. When the computer program is executed in a computer, the computer is caused to execute the above-mentioned hand hygiene management method based on image data.

[0013] Other features and advantages will be described in the following description. The objectives and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description and drawings.

[0014] 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

[0015] 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 accompanying drawings.

[0016] Figure 1 It is a flow chart of a hand hygiene management method based on image data; Figure 2 is a flow chart of a method for dividing a video stream into groups of similar video streams; Figure 3 is a flow chart of a method for determining a target type of operating room; Figure 4 The present invention is a flowchart of a method for determining a sensitivity abnormal combination among similar video stream combinations. DETAILED DESCRIPTION

[0017] To help those skilled in the art better understand the technical solutions in this specification, the following will provide a clear and complete description of the technical solutions in the embodiments of this specification, in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of this specification, not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this specification without creative work should fall within the scope of protection of this specification.

[0018] In the present application, the recognition results of the video frames in the video stream used to detect the cleaning results of hand hygiene are divided into different similar video stream combinations, and when there are many similar video stream combinations, a verification image model is constructed to ensure that the reliability of the recognition processing of the cleaning status of hand hygiene is improved when the cleaning posture is inconsistent.

[0019] Example 1 like Figure 1 As shown, the present application provides a hand hygiene management method based on image data, specifically including: S1 determines similarities among video frames of a hand hygiene video stream based on the analysis results of the hand hygiene monitoring data, divides the video stream into similar video stream combinations based on the similarities, and proceeds to the next step when it is determined that the target type of operating room meets the requirements based on the video stream data of each similar video stream combination in the monitoring image of the operating room cleaning device; Furthermore, the similarity of the video frames is determined based on the image similarity of different video frames in the video stream, specifically based on the similarity of texture features of the video frames, specifically by calculating the texture feature similarity using a Euclidean distance function.

[0020] Specifically, the video stream is a monitoring video of each cleaning process as a single video stream.

[0021] Specifically, such as Figure 2 As shown, the video stream is divided into various similar video stream combinations, specifically including: Based on similarities between the video stream and other video streams in different video frames, determining video frames of the video stream having similar video frames in other video streams as similar video frames; taking a video frame similar to the similar video frame in the other video stream as a corresponding video frame; According to similar video frame data of the video stream and corresponding video frame data in other video streams, it is determined whether the video stream and other video streams can be classified into the same similar video stream combination.

[0022] Specifically, the similar video frame of the video stream is a video frame of the video stream whose texture feature similarity meets the requirements in other video streams. Specifically, when the video frame has a texture feature similarity greater than 0.8 in other video streams, the video frame is regarded as a similar video frame.

[0023] It can be understood that the corresponding video frames in the other video streams are video frames in the other video streams whose texture feature similarity with similar video frames in the video stream meets the requirement.

[0024] In a possible embodiment, if the number of similar video frames in the video stream meets the requirements and the number of corresponding video frames of different similar video frames in the other video frames after deduplication processing meets the requirements, the video stream is divided into the same similar video stream combination. It can be understood that when the number of similar video frames in the video stream accounts for more than 90% and the number of corresponding video frames of different similar video frames in the other video frames after deduplication processing accounts for more than 80%, the video stream is divided into the same similar video stream combination.

[0025] The deduplication process is to perform deduplication on the same object video frame among different similar video frames, and only treat the corresponding video frames of multiple similar video frames as one corresponding video frame.

[0026] Optionally, dividing the video stream into similar video stream combinations specifically includes: Based on similarities between the video stream and other video streams in different video frames, determining video frames of the video stream having similar video frames in other video streams as similar video frames; Determine similar video frames in the other video streams by using similarities between the video frames in the other video streams and the video frames in the video stream; According to similar video frame data in the video stream and other video streams, it is determined whether the video stream and other video streams can be divided into the same similar video stream combination.

[0027] It can be understood that when the similar video frame data in the video stream and other video streams meet the requirements, that is, the number of similar video frames in the video stream is greater than 90% and the number of similar video frames in other video streams accounts for more than 80%, it is determined that the video stream and other video streams can be divided into the same similar video stream combination.

[0028] Specifically, such as Figure 3 As shown, the method for determining the target type of operating room is: Determining the number of video streams of each similar video stream combination based on video stream data of each similar video stream combination in history of the monitoring image of the cleaning device of the operating room; Whether the operating room is an operating room of the target type is determined based on the number of video streams combined with each video stream.

[0029] It should be noted that when the number of similar video stream combinations whose proportion in the video streams in the monitoring image of the cleaning device is greater than 10% is more than 4, there are deviations in the hand hygiene cleaning methods of different personnel. Therefore, when a single image recognition model is used on this basis and there is an identification deviation in any similar video stream combination, it will affect the reliability of the hand hygiene monitoring and processing in the operating room, and the operating room is determined to be the target type of operating room.

[0030] S2 determines the abnormal alarm data within each cleaning time interval based on the video stream data in the similar video stream combination, determines the sensitivity abnormal combination in the similar video stream combination based on the abnormal alarm data, and uses the video stream data of the sensitivity abnormal combination in the monitoring image of the cleaning device in each operating room to determine whether it is necessary to construct a verification image recognition model. The video stream data of the sensitivity abnormal combination in the monitoring image of the cleaning device in the operating room is used to determine whether the verification image recognition model needs to perform verification processing on the monitoring image of the similar video stream combination in the cleaning device of the operating room.

[0031] It is understandable that the cleaning time intervals are divided into equal intervals, specifically, the cleaning time intervals are divided into intervals of 15 seconds.

[0032] Furthermore, the abnormal alarm data within the cleaning time interval includes the number of abnormal alarms within the cleaning time interval, specifically the number of alarms in the history due to identification of hand hygiene cleaning abnormalities.

[0033] Specifically, such as Figure 4 As shown, the method for determining the sensitivity abnormal combination in the similar video stream combination is: Based on the abnormal alarm data, determining the number of abnormal alarms of the similar video stream combination within different cleaning time intervals in history; Using the number of abnormal alarms, determining abnormal alarm rates within different cleaning time intervals; Whether the similar video stream combination is a sensitivity abnormality combination is determined according to the abnormal alarm rate in different cleaning time intervals.

[0034] It should be noted that the abnormal alarm rate is determined by the ratio of the number of abnormal alarms of the video streams of the similar video stream combination within the cleaning time interval to the number of video streams of the similar video stream combination within the cleaning time interval.

[0035] It is understandable that determining whether the similar video stream combination is a sensitivity abnormal combination based on the abnormal alarm rate in different cleaning time intervals specifically includes: If the deviation amount of the abnormal alarm rate within the cleaning time interval shorter than the cleaning time interval does not meet the requirement and the number of cleaning time intervals does not meet the requirement, the similar video stream combination is determined to be a sensitivity abnormality combination.

[0036] It can be understood that the cleaning time interval with a shorter cleaning time is a cleaning time interval that is smaller than the maximum value of the endpoints of the cleaning time interval.

[0037] Among them, if in the similar video stream combination, the number of cleaning time intervals with an abnormal alarm rate greater than 10% within the cleaning time interval shorter than the cleaning time is not less than 2, that is, when the abnormal alarm rate of the cleaning time interval with a larger maximum endpoint value is greater than 10% of the cleaning time interval with a smaller maximum endpoint value, and at this time, the number of cleaning time intervals with a larger maximum endpoint value is not less than 2, then the similar video stream combination is determined to be a sensitivity abnormal combination. At this time, as the cleaning time increases, the abnormal alarm rate will increase instead. Therefore, the similar video stream combination can be regarded as a sensitivity abnormal combination.

[0038] In another possible embodiment, the number of video streams in the cleaning time interval with a shorter cleaning time and the cleaning time interval with a higher abnormality alarm rate than the cleaning time interval with a shorter cleaning time must both be more than 100 times.

[0039] Furthermore, it is determined that a verification image recognition model needs to be constructed, including: Based on the video stream data in the monitoring images of the cleaning devices in each operating room, determining the number of video streams with abnormal sensitivity combinations in the monitoring images of the cleaning devices in the operating room in history; Determine, based on the number of the video streams, an operating room having a video stream with an abnormal sensitivity combination, and use it as an operating room with abnormal sensitivity; The number of video streams with abnormal sensitivity combinations in the abnormal sensitivity operating room is used to determine whether it is necessary to build a verification image recognition model.

[0040] It can be understood that when the proportion of the number of video streams with abnormal sensitivity combinations does not meet the requirement of a large number of operating rooms with abnormal sensitivity, it is determined that a verification image recognition model needs to be constructed, wherein when the number of video streams with abnormal sensitivity combinations accounts for more than 30% and the number of operating rooms with abnormal sensitivity is more than 2, it is determined that a verification image recognition model needs to be constructed, wherein the proportion of the number of video streams with abnormal sensitivity combinations to the number of video streams in the monitoring image of the operating room with abnormal sensitivity is used as the proportion of the number of video streams with abnormal sensitivity combinations.

[0041] Further, determining whether verification processing of the monitoring image of the similar video stream combination is required specifically includes: The operating room containing the similar video stream combination in the monitoring image of the cleaning device is used as the monitoring target operating room, and the number of video streams with abnormal sensitivity combinations in different monitoring target operating rooms is determined by using the historical monitoring data of the monitoring images of different monitoring target operating rooms, and the optional operating room in the monitoring target operating room is determined by using the number of video streams; Determining the number of video streams of the different optional operating rooms in the similar video stream combination according to historical monitoring data of monitoring images of the different optional operating rooms; Based on the number of video streams with abnormal sensitivity combinations and the number of video streams with similar video stream combinations in different optional operating rooms, determine the monitoring images of the video streams of the similar video stream combinations in the cleaning device of the operating room, and whether verification processing of the monitoring images of the similar video stream combinations is required.

[0042] It can be understood that if the ratio of the number of video streams in the similar video stream combination of optional operating rooms to the total number of video streams in the similar video stream combination is less than a preset threshold or the number of optional operating rooms is less than the threshold for the number of optional operating rooms, then all the remaining optional operating rooms are determined to be the operating rooms for verification processing of the monitoring images of the similar video stream combination.

[0043] Furthermore, if the ratio of the number of video streams in the similar video stream combination of the optional operating room to the total number of video streams of the similar video stream combination is not less than a preset threshold and the number of optional operating rooms is not less than the threshold of the number of optional operating rooms, the matching value is determined based on the proportion of the video streams of the similar video stream combination in the optional operating room and the proportion of the video streams of the sensitivity abnormality combination, and a preset number of operating rooms for verification processing of the monitoring images of the similar video stream combination are selected from large to small according to the matching value, wherein the matching value is obtained by subtracting the sum of 1 and the proportion of the number of video streams of the similar video stream combination in the optional operating room and the proportion of the number of video streams of the sensitivity abnormality combination, and its value range is between 0 and 1.

[0044] Among them, if the video streams of a preset number of similar video stream combinations account for less than 50% of the video streams of the similar video stream combination in the history, and if the number of video streams of the similar video stream combination in the current date appears at least 2 times, the monitoring image of the similar video stream combination will be verified the next time it appears, thereby ensuring the reliability of the verification process of hand hygiene in the operating room; in other cases, if the number of video streams of the similar video stream combination in the current date appears more than 3 times, the monitoring image of the similar video stream combination will be verified the next time it appears, thereby ensuring the reliability of the verification process of hand hygiene in the operating room.

[0045] It should also be noted that if the number of inconsistencies during the verification process of the monitoring images of the similar video stream combination is greater than 5%, the verification process of the monitoring images of the similar video stream combination is performed in all operating rooms.

[0046] Example 2 In a second aspect, the present invention provides a computer system comprising: a memory and a processor in communication connection, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the above-mentioned hand hygiene management method based on image data when running the computer program.

[0047] Optionally, the method for determining the target type of operating room is: Determining the number of video streams of each similar video stream combination based on video stream data of each similar video stream combination in history of the monitoring image of the cleaning device of the operating room; Determine the proportion of the number of video streams in each video stream combination in the video stream in the monitoring image of the cleaning device based on the number of video streams in each video stream combination; Whether the operating room is an operating room of the target type is determined according to the proportion of the number of video streams in the video streams in the monitoring image of the cleaning device.

[0048] It can be understood that when there is no similar video stream combination in which the number of video streams accounts for more than 80% of the video streams in the monitoring image of the cleaning device, the operating room is an operating room of the target type.

[0049] Specifically, when there are a large number of operating rooms of the target type, the recognition reliability of a single image recognition model is low, so it is determined that the target type of operating rooms is insufficient. At this time, it is necessary to construct a verification image recognition model. Since there are a large number of operating rooms of the target type, it is necessary to verify the video stream of the operating room of the target type. Only when the video stream of other operating rooms is not in the similar video stream combination of the operating room of the target type, it is necessary to verify the video stream of the cleaning device of other operating rooms.

[0050] It should be noted that the cleaning device of the operating room is a faucet for hand hygiene cleaning before entering the operating room, specifically the faucet corresponding to the operating room. Different operating rooms have corresponding independent faucets before entering the operating room.

[0051] Optionally, determine whether to build a verification image recognition model, including: Based on the video stream data in the monitoring images of the cleaning devices in each operating room, determining the number of video streams with abnormal sensitivity combinations in the monitoring images of the cleaning devices in the operating room in history; Determine, based on the number of the video streams, an operating room having a video stream with an abnormal sensitivity combination, and use it as an operating room with abnormal sensitivity; The need to construct a verification image recognition model is determined based on the proportion of video streams with abnormal sensitivity combinations in the abnormal sensitivity operating room.

[0052] It should be noted that when there is a sensitivity abnormality operating room in which the number of video streams with sensitivity abnormality combinations accounts for a greater proportion than the preset value of the abnormality proportion, it is determined that the construction of the verification image recognition model needs to be carried out.

[0053] It is understandable that when there is no need to construct a verification image recognition model, there is no need to use the verification image recognition model for verification processing, and only a single image recognition model is needed to perform recognition processing of hand hygiene in the monitoring image.

[0054] Example 3 In a third aspect, the present invention provides a computer storage medium having a computer program stored thereon. When the computer program is executed in a computer, the computer is caused to execute the above-mentioned hand hygiene management method based on image data.

[0055] Optionally, determining whether verification processing of the monitoring image of the similar video stream combination is required specifically includes: The operating room whose monitoring image of the cleaning device has the similar video stream combination in history is used as the monitoring target operating room, and the number of video streams of the different monitoring target operating rooms in the similar video stream combination is determined based on the historical monitoring data of the monitoring images of the different monitoring target operating rooms; Using historical monitoring data of monitoring images of different monitoring target operating rooms, determining the number of video streams with abnormal sensitivity combinations in different monitoring target operating rooms; Based on the number of video streams with abnormal sensitivity combinations and the number of video streams with similar video stream combinations in different monitoring target operating rooms, the monitoring images of the video streams of the similar video stream combinations in the cleaning device of the operating room are determined, and whether verification processing of the monitoring images of the similar video stream combinations is required.

[0056] It should be noted that when the proportion of the number of video streams with abnormal sensitivity combinations in the monitored target operating room meets the requirements, that is, the proportion of the number of video streams with abnormal sensitivity combinations in the monitored target operating room is too high, that is, the proportion of the number of video streams with abnormal sensitivity combinations in the monitored target operating room is greater than 50%, at this time, since the verification processing of the monitoring images of the video streams with abnormal sensitivity combinations can ensure the monitoring reliability of the monitored target operating room, on this basis, it is determined that there is no need to perform verification processing of the monitoring images of the similar video stream combination in the monitored target operating room.

[0057] Furthermore, the monitoring target operating rooms whose proportion of video streams of the sensitivity abnormal combination meets the requirements are removed as optional operating rooms. If the ratio of the number of video streams of the optional operating room in the similar video stream combination to the total number of video streams of the similar video stream combination is less than a preset threshold or the number of optional operating rooms is less than the threshold for the number of optional operating rooms, all the remaining optional operating rooms are determined as operating rooms for verification processing of the monitoring images of the similar video stream combination.

[0058] In a possible embodiment, if the ratio of the number of video streams in the similar video stream combination of optional operating rooms to the total number of video streams of the similar video stream combination is less than 60% or the number of optional operating rooms is less than 4, then all the remaining optional operating rooms will be used as operating rooms for verification processing of the monitoring images of the similar video stream combination. It should also be noted that at this time, due to the poor monitoring reliability, as long as the number of video streams that do not belong to the sensitivity abnormality combination meets the requirements of the monitoring target operating room, as long as the video stream of the similar video stream combination appears, the verification processing of the monitoring image of the similar video stream combination will be performed.

[0059] It can also be understood that when the proportion of the number of video streams of the sensitivity abnormality combination in the monitored target operating room meets the requirements, and the ratio of the number of video streams of the similar video stream combination in the optional operating room to the total number of video streams of the similar video stream combination is not less than the preset threshold or the number of optional operating rooms is not less than the threshold of the number of optional operating rooms, it is necessary to further determine whether the proportion of the number of video streams of the similar video stream combination in the optional operating room meets the requirements. When the proportion of the number of video streams of the similar video stream combination in the optional operating room meets the requirements, that is, the proportion of the video streams of the similar video stream combination in the optional operating room in the video streams in the optional operating room is greater than 20%, it is determined to perform verification processing of the monitoring image of the similar video stream combination in the optional operating room.

[0060] If the ratio of the number of video streams of the similar video stream combination in the optional operating room does not meet the requirements, then in the optional operating rooms excluding the operating rooms where the verification processing of the monitoring images of the similar video stream combination is performed and the monitoring target operating rooms where the ratio of the number of video streams of the sensitivity abnormal combination meets the requirements, the optional operating rooms for which the verification processing of the monitoring images of the similar video stream combination is performed are determined according to the ratio of the number of video streams of the sensitivity combination from small to large, until a preset number of optional operating rooms excluding the optional operating rooms where the verification processing of the monitoring images of the similar video stream combination is performed, that is, 3 optional operating rooms are selected for the verification processing of the monitoring images of the similar video stream combination. On this basis, if in the operating rooms that do not belong to the operating rooms where the verification processing of the monitoring images of the similar video stream combination is performed and the operating rooms that do not belong to the monitoring target operating rooms where the ratio of the number of video streams of the sensitivity abnormal combination meets the requirements, if the number of video streams of the similar video stream combination appears more than 3 times in the current date, the verification processing of the monitoring image of the similar video stream combination will be performed the next time it appears, thereby ensuring the reliability of the verification processing of the hand hygiene in the operating room.

[0061] It should also be noted that if the number of inconsistencies during the verification process of the monitoring images of the similar video stream combination is greater than 5%, the verification process of the monitoring images of the similar video stream combination is performed in all operating rooms.

[0062] The various embodiments in this specification are described in a progressive manner. Similar portions between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from the other embodiments. In particular, the device, apparatus, and non-volatile computer storage medium embodiments are generally similar to the method embodiments, so their descriptions are relatively simplified. For relevant details, refer to the descriptions of the method embodiments.

[0063] The foregoing description of this specification describes specific embodiments. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that described in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order shown or the sequential order to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0064] The foregoing description is merely one or more embodiments of this specification and is not intended to limit this specification. It will be apparent to those skilled in the art that various modifications and variations may be made to one or more embodiments of this specification. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of one or more embodiments of this specification are intended to be within the scope of the claims of this specification.

Claims

1. A hand hygiene management method based on image data, characterized in that: Specifically include: The hand hygiene monitoring data is analyzed to determine similarities among video frames of the hand hygiene video stream. Based on the similarities, the video stream is divided into similar video stream combinations. When it is determined that the target type of operating room meets the requirements based on the video stream data of each similar video stream combination in the monitoring image of the operating room cleaning device, the next step is performed. Based on the video stream data in the similar video stream combination, the abnormal alarm data in each cleaning time interval is determined, and based on the abnormal alarm data, the sensitivity abnormal combination in the similar video stream combination is determined. When the video stream data of the sensitivity abnormal combination in the monitoring image of the cleaning device in each operating room is used to determine whether it is necessary to construct a verification image recognition model, the video stream data of the sensitivity abnormal combination in the monitoring image of the cleaning device in the operating room is used to determine whether the verification image recognition model needs to perform verification processing of the monitoring image of the similar video stream combination in the cleaning device of the operating room.

2. The hand hygiene management method based on image data according to claim 1, characterized in that: The similarity of the video frames is determined based on the degree of image similarity between different video frames in the video stream.

3. The hand hygiene management method based on image data according to claim 1, characterized in that: The video stream is to use the monitoring video of each cleaning process as a single video stream.

4. The hand hygiene management method based on image data according to claim 1, characterized in that: Dividing the video stream into similar video stream combinations, specifically including: Based on similarities between the video stream and other video streams in different video frames, determining video frames of the video stream having similar video frames in other video streams as similar video frames; taking a video frame similar to the similar video frame in the other video stream as a corresponding video frame; According to similar video frame data of the video stream and corresponding video frame data in other video streams, it is determined whether the video stream and other video streams can be classified into the same similar video stream combination.

5. The hand hygiene management method based on image data according to claim 4, characterized in that: The similar video frames of the video stream are video frames of the video stream that have texture feature similarity that meets the requirement in other video streams.

6. The hand hygiene management method based on image data according to claim 1, characterized in that: The abnormal alarm data within the cleaning time interval includes the number of abnormal alarms within the cleaning time interval.

7. The hand hygiene management method based on image data according to claim 1, characterized in that: Determine the need to build a verification image recognition model, including: Based on the video stream data in the monitoring images of the cleaning devices in each operating room, determining the number of video streams with abnormal sensitivity combinations in the monitoring images of the cleaning devices in the operating room in history; Determine, based on the number of the video streams, an operating room having a video stream with an abnormal sensitivity combination, and use it as an operating room with abnormal sensitivity; The number of video streams with abnormal sensitivity combinations in the abnormal sensitivity operating room is used to determine whether it is necessary to build a verification image recognition model.

8. The hand hygiene management method based on image data according to claim 1, characterized in that: Determining whether verification processing of the monitoring image of the similar video stream combination is required specifically includes: The operating room whose monitoring image of the cleaning device has the similar video stream combination in history is used as the monitoring target operating room, and the number of video streams of the different monitoring target operating rooms in the similar video stream combination is determined based on the historical monitoring data of the monitoring images of the different monitoring target operating rooms; Using historical monitoring data of monitoring images of different monitoring target operating rooms, determining the number of video streams with abnormal sensitivity combinations in different monitoring target operating rooms; Based on the number of video streams with abnormal sensitivity combinations and the number of video streams with similar video stream combinations in different monitoring target operating rooms, the monitoring images of the video streams of the similar video stream combinations in the cleaning device of the operating room are determined, and whether verification processing of the monitoring images of the similar video stream combinations is required.

9. A computer system comprising: A memory and a processor in communication connection, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the hand hygiene management method based on image data according to any one of claims 1 to 8 when running the computer program.

10. A computer storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to execute the hand hygiene management method based on image data according to any one of claims 1 to 8.

Images