Video data processing method and device, equipment, storage medium and program product

Abstract

The invention relates to a video data processing method and device, a storage medium and a program product, and the method comprises the steps: obtaining a to-be-processed video, and determining the initial structured information of the to-be-processed video according to the to-be-processed video; the initial structured information comprises to-be-processed label set information associated with the to-be-processed video; verifying the initial structured information according to predefined structured information to obtain target structured information of the to-be-processed video; the predefined structured information includes tag set information associated with the surgical video stored in the database. By using the video data processing method provided by the invention, the accuracy of the obtained target structured information can be improved.

Description

Technical Field

[0001] The present application relates to the field of medical technology, and in particular to a video data processing method, apparatus, device, storage medium and program product. Background Art

[0002] Minimally invasive surgery, which is widely used nowadays, generates a huge amount of video data, each of which contains a large amount of structured information associated with the surgical video. Therefore, it is very important to perform video structuring processing on each video to obtain the structured information corresponding to each video.

[0003] In traditional technology, a convolutional neural network is used to perform video structuring processing to obtain structured information corresponding to the video.

[0004] However, the above-mentioned video structuring processing method has the problem of low accuracy of the obtained structured information. Summary of the invention

[0005] Based on this, it is necessary to provide a video data processing method, device, equipment, storage medium and program product that can improve the accuracy of the acquired structured information in response to the above technical problems.

[0006] In a first aspect, the present application provides a video data processing method, the method comprising:

[0007] Get the video to be processed;

[0008] According to the video to be processed, initial structured information of the video to be processed is determined; the initial structured information includes information of a tag set to be processed associated with the video to be processed;

[0009] The initial structured information is verified according to the predefined structured information to obtain the target structured information of the video to be processed; wherein the predefined structured information includes the label set information associated with the surgical video stored in the database.

[0010] In one embodiment, determining initial structured information of the video to be processed according to the video to be processed includes:

[0011] Extracting multiple video frame images from the video to be processed according to a preset extraction rule;

[0012] A plurality of video frame images are input into a first detection model to obtain initial structural information of the video to be processed.

[0013] In one embodiment, the method further comprises:

[0014] Identify the region of interest in each video frame image to obtain the region of interest image corresponding to each video frame image;

[0015] A plurality of region of interest images are input into a first detection model to obtain initial structural information of the video to be processed.

[0016] In one embodiment, the initial structured information includes image-level information and time-series-level information, the first detection model includes an image-level detection model and a time-series-level detection model, and multiple video frame images are input into the first detection model to obtain the initial structured information of the video to be processed, including:

[0017] Input each video frame image into the image-level detection model to obtain the image-level information corresponding to each video frame image;

[0018] The multiple video frame images are divided into multiple video frame sets according to the time sequence, and each video frame set is sequentially input into the time sequence level detection model to obtain the time sequence level information corresponding to each video frame set.

[0019] In one embodiment, the initial structured information is verified according to the predefined structured information to obtain the target structured information of the video to be processed, including:

[0020] Searching for matching structured information corresponding to the initial structured information from predefined structured information; the similarity between the matching structured information and the initial structured information is greater than a preset similarity threshold;

[0021] The initial structured information is simplified according to the matching structured information to obtain the target structured information.

[0022] In one embodiment, verifying the initial structured information according to the matching structured information to obtain the target structured information includes:

[0023] Comparing the initial structured information with the matching structured information to obtain a comparison result; the comparison result includes information indicating that abnormal labels exist in the initial structured information;

[0024] The initial structured information is verified according to the comparison results to obtain the target structured information.

[0025] In one embodiment, the initial structured information is verified according to the predefined structured information to obtain the target structured information of the video to be processed, including:

[0026] Determining associated structured information corresponding to the initial structured information in the predefined structured information;

[0027] Setting a ratio mark for other structured information except the associated structured information in the predefined structured information; the ratio mark is used to represent the probability corresponding to the other structured information;

[0028] The initial structured information is detected according to the ratio mark, the associated structured information and other structured information to obtain the target structured information.

[0029] In one embodiment, a plurality of video frame images are input into a first detection model to obtain initial structured information of the video to be processed, including:

[0030] According to the original video and the corresponding structured information, the first detection model is trained to obtain a second detection model;

[0031] A plurality of video frame images are input into the second detection model to obtain initial structured information of the video to be processed.

[0032] In one embodiment, the method further comprises:

[0033] The target structured information is corrected by using a time-series correction method to obtain the corrected target structured information.

[0034] In one embodiment, the target structured information is corrected by using a time-series correction method to obtain the corrected target structured information, including:

[0035] Performing temporal smoothing processing on target structure information corresponding to multiple video frame images to obtain corrected target structure information;

[0036] and / or;

[0037] According to the correlation between the target structured information corresponding to each video frame image, the target structured information is corrected to obtain the corrected target structured information.

[0038] In one embodiment, the method further comprises:

[0039] A correspondence between the target structured information and the video to be processed is established, and the correspondence is stored.

[0040] In a second aspect, an embodiment of the present application provides a video data processing device, the device comprising:

[0041] An acquisition module is used to acquire the video to be processed;

[0042] A determination module, configured to determine initial structured information of the video to be processed according to the video to be processed; the initial structured information includes information of a tag set to be processed associated with the video to be processed;

[0043] A verification module is used to verify the initial structured information according to predefined structured information to obtain target structured information of the video to be processed; wherein the predefined structured information includes label set information associated with the surgical video stored in a database.

[0044] In a third aspect, an embodiment of the present application provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method provided in the first aspect when executing the computer program.

[0045] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method provided in the first aspect when the computer program is executed by a processor.

[0046] In a fifth aspect, an embodiment of the present application further provides a computer program product, including a computer program, which implements the steps of the method provided in the first aspect when the computer program is executed by a processor.

[0047] The above-mentioned video data processing method, device, equipment, storage medium and program product, the method obtains the video to be processed; determines the initial structured information of the video to be processed according to the video to be processed; the initial structured information includes the tag set information to be processed associated with the video to be processed; verifies the initial structured information according to the predefined structured information to obtain the target structured information of the video to be processed; the predefined structured information includes the tag set information associated with the surgical video stored in the database. In this embodiment, the video to be processed is first detected to obtain the initial structured information of the video to be processed; and then the initial structured information is verified according to the predefined structured information, that is, after the initial structured information is obtained, it will be determined whether it is accurate according to the predefined structured information, and it will be modified if it is inaccurate, so that the target structured information corresponding to the obtained video to be processed is richer and more accurate, that is, it can improve the accuracy of the target structured information corresponding to the obtained video to be processed. In addition, the present application directly uses the tag set information associated with the surgical video stored in the database to verify the tag set information to be processed associated with the video to be processed in the initial structured information, so that the efficiency of verifying the initial structured information can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 A diagram showing an application environment of a video data processing method in an embodiment;

[0049] Figure 2 A schematic diagram of a process flow of a video data processing method in one embodiment;

[0050] Figure 3A schematic diagram of a step flow chart of a video data processing method in another embodiment;

[0051] Figure 4 A schematic diagram of a step flow chart of a video data processing method in another embodiment;

[0052] Figure 5 A schematic diagram of a step flow chart of a video data processing method in another embodiment;

[0053] Figure 6 A schematic diagram of a step flow chart of a video data processing method in another embodiment;

[0054] Figure 7 A schematic diagram of a step flow chart of a video data processing method in another embodiment;

[0055] Figure 8 A schematic diagram of a step flow chart of a video data processing method in another embodiment;

[0056] Fig. 9 A schematic diagram of a step flow chart of a video data processing method in another embodiment;

[0057] Fig.10 A schematic diagram of a step flow chart of a video data processing method in another embodiment;

[0058] Fig.11 A schematic diagram of a step flow chart of a video data processing method in another embodiment;

[0059] Fig.12 is a schematic diagram of target structured information in one embodiment;

[0060] Fig.13 FIG. 4 is a schematic diagram of the structure of a video data processing device in an embodiment. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0062] The serial numbers assigned to the components in this application, such as "first", "second", etc., are only used to distinguish the objects described and do not have any order or technical meaning.

[0063] First of all, before specifically introducing the technical solutions of the embodiments disclosed in the present application, the background technology or technical evolution context on which the embodiments of the present application are based is introduced. With the development of medical technology, the currently widely used minimally invasive surgery will generate a large amount of video data, and each video data includes a large amount of structured information. Therefore, it is very important to perform video structuring processing on each video to obtain the structured information corresponding to each video. However, due to the large amount of video data and the long length of a single video, it is difficult to perform quantitative analysis on the video data. In the traditional technology, a convolutional neural network is used to perform video structuring processing on the video, and the structured information corresponding to the obtained video is less and inaccurate, resulting in a large amount of video data not being utilized, resulting in a waste of data. In this regard, the present application provides a video data processing method that can obtain more accurate structured information.

[0064] The technical solution of the present application and how the technical solution of the present application solves the technical problem are described in detail below with specific embodiments.

[0065] The video data processing method provided in the present application can be applied to computer devices, which may be, but are not limited to, industrial computers, laptops, and tablet computers. The internal structure diagram of the computer device can be as follows: Figure 1 As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through 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 a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a video data processing method is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a key, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0066] In one embodiment, Figure 2 As shown, a video data processing method is provided. This embodiment takes the method applied to a computer device as an example for illustration. In this embodiment, the method includes the following steps:

[0067] Step 200: Obtain the video to be processed.

[0068] The video to be processed is a video for which corresponding structured information needs to be obtained. The video to be processed can be a video stored in a database, or a video obtained in real time during a minimally invasive surgery. The computer device can obtain the video to be processed from the database, or can communicate with the corresponding system of the hospital to obtain the video to be processed of the surgical process in real time. This embodiment does not limit the specific method of obtaining the video to be processed, as long as its function can be achieved.

[0069] In an optional embodiment, when the computer device needs to determine the target structured information corresponding to the video to be processed, it is also necessary to obtain predefined structured information. The predefined structured information may be a tag set information associated with a large number of surgical videos pre-stored in a database of the computer device or the corresponding hospital system. The predefined structured information may include information manually input into the database by the user based on the surgical video, that is, part of the publicly available information such as the patient, doctor and hospital corresponding to the surgical video; and the computer device analyzes the surgical video to obtain the tag set information. This embodiment does not limit the specific content of the predefined structured information stored in the database, and the specific method of obtaining the predefined structured information. Patient information may include but is not limited to name, gender, age and BMI (Body Mass Index, body mass index), etc. Doctor information includes but is not limited to participants corresponding to historical surgical videos, the rank of each participant, etc. Hospital information may include but is not limited to hospital name, affiliated level, etc. The tag set information corresponding to the surgical video may include but is not limited to surgical procedure, surgical image quality, instrument category, anatomical structure, surgical stage, procedure-instrument category, surgical abnormal events, etc. The predefined structured information in the database may be stored in the form of indexes (tags). An optional storage instance is shown in the following table:

[0070]

[0071] Among them, Short (short integer type), Long (long integer type) or Tuple (array type) is the type of structured information storage. Calot triangle dissection indicates the stage of removing the connection between the gallbladder and the liver; clipping indicates the stage of removing the cystic artery.

[0072] Step 210: Determine initial structured information of the video to be processed according to the video to be processed; the initial structured information includes information of a tag set to be processed associated with the video to be processed.

[0073] After acquiring the video to be processed, the computer device analyzes and processes the video to be processed, and the initial structured information of the video to be processed can be obtained. That is to say, by analyzing the initial video, the information associated with the surgery contained in the video to be processed, that is, the label set information to be processed can be obtained. The label information to be processed may include the patient information, doctor information, hospital information corresponding to the video to be processed, as well as the image quality of each frame of the video to be processed, the surgical procedure included in each frame of the video, the type of equipment, and other information. Among them, the patient information, doctor information, and hospital information can be the information associated with the video to be processed that is directly stored when the video to be processed is stored. The specific method for detecting and determining the initial structured information of the video to be processed in this embodiment is not limited, as long as its function can be achieved.

[0074] Step 220: verify the initial structured information according to the predefined structured information to obtain the target structured information corresponding to the video to be processed; wherein the predefined structured information includes the label set information associated with the surgical video stored in the database.

[0075] After obtaining the initial structured information of the video to be processed, the computer device verifies the initial structured information according to the obtained predefined structured information, and can obtain more accurate target structured information corresponding to the video to be processed.

[0076] The predefined structured information contains a large amount of accurate and rich structured information corresponding to the surgical videos. The computer device can search for relevant structured information in the predefined structured information based on the initial structured information obtained by detecting and processing the video to be processed, and modify the abnormal label information in the initial structured information of the video to be processed based on the relevant structured information, so as to obtain more accurate and rich target structured information corresponding to the video to be processed. This embodiment does not limit the specific method of verifying the initial structured information according to the predefined structured information, as long as its function can be achieved.

[0077] The video data processing method provided in the embodiment of the present application is obtained by obtaining the video to be processed; determining the initial structured information of the video to be processed according to the video to be processed; the initial structured information includes the tag set information to be processed associated with the video to be processed; the initial structured information is verified according to the predefined structured information to obtain the target structured information of the video to be processed; the predefined structured information includes the tag set information associated with the surgical video stored in the database. In this embodiment, the video to be processed is first detected to obtain the initial structured information of the video to be processed; and then the initial structured information is verified according to the predefined structured information, that is, after the initial structured information is obtained, it will be determined whether it is accurate according to the predefined structured information, and it will be modified if it is inaccurate, so that the target hand structured information corresponding to the video to be processed is obtained. More abundant and accurate, that is, the accuracy of the target structured information corresponding to the obtained video to be processed can be improved. In addition, the present application directly uses the tag set information associated with the surgical video stored in the database to verify the tag set information to be output associated with the video to be processed in the initial structured information, which can improve the efficiency of verifying the initial structured information.

[0078] In one embodiment, Figure 3 As shown, it is related to an implementation method of determining the initial structured information of the video to be processed according to the video to be processed, and the steps of the implementation method include:

[0079] Step 300: extract multiple video frame images from the video to be processed according to preset extraction rules.

[0080] When obtaining the video to be processed, the computer device divides the video to be processed into a plurality of video frame images. Specifically, the computer device may extract a plurality of video frame images from the video to be processed according to a preset extraction rule. The preset extraction rule may be pre-stored in the computer device by a user. The preset extraction rule may be to extract each frame image in the video to be processed; or it may be to extract a plurality of frames of images in the video to be processed at a preset interval, for example, 2 frames per second. This embodiment does not limit the preset extraction rule, and the user may set it according to actual needs.

[0081] Step 310: Input multiple video frame images into a first detection model to obtain initial structured information of the video to be processed.

[0082] The first detection model may be pre-stored in a computer device. The first detection model is trained using a plurality of video frame image samples corresponding to a surgical video sample and a structured information sample corresponding to the surgical video sample. This embodiment does not limit the specific process of training the first detection model, as long as its function can be achieved.

[0083] After obtaining multiple video frame images corresponding to the video to be processed, the computer device inputs the multiple video frame images into the first detection model, and detects them through the first detection model to obtain the initial structural information of the video to be processed. The computer device can input each video frame image into the first detection model for detection in sequence, or divide the multiple video frame images into multiple video frame sets and input them into the first detection model respectively. The specific input method can be set according to the input method in the training process of the first detection model.

[0084] In this embodiment, multiple video frame images extracted from the video to be processed are directly input into the pre-trained first detection model to obtain the initial structured information of the video to be processed. This method of obtaining the initial structured information is simple and fast, and can improve the efficiency of determining the target structured information corresponding to the video to be processed.

[0085] In one embodiment, after obtaining multiple video frame images in the video to be processed, the computer device may first identify and analyze each video frame image to obtain the corresponding region of interest, and then detect it through the first detection model to obtain the initial structured information of the video to be processed. Figure 4 As shown, the specific steps include:

[0086] Step 400: Identify the region of interest in each video frame image to obtain the region of interest image corresponding to each video frame image.

[0087] After obtaining multiple video frame images corresponding to the video to be processed, the computer device performs recognition processing on each video frame image to obtain an image of the region of interest in each video frame image. This embodiment does not limit the specific method for identifying the region of interest in each video frame image, as long as its function can be achieved. Assuming that the video to be processed is a minimally invasive surgery video, the region of interest in each video frame image refers to the endoscope area.

[0088] In an optional embodiment, the computer device may sequentially input each video frame image into a segmentation model pre-trained by deep learning, and segment the region of interest image in each video frame image by the segmentation model. The computer device may also use machine learning to model the region of interest and other regions, and use the model to segment the region of interest in each video frame image to obtain the region of interest image.

[0089] In another optional embodiment, for each video frame image, the computer device may first grayscale the video frame image to obtain a grayscale image; binarize the grayscale image to obtain a binary image; then perform a connected domain analysis on the binary image to determine the maximum connected domain; and cut out the maximum connected domain in the binary image to obtain an image of the region of interest. In other words, the maximum connected domain is the region of interest. Among them, the method for grayscale processing of the video frame image may include any one of the R (red) G (green) B (blue) average method, the RGB weighted average method, the H (hue) I (brightness) S (saturation) conversion method, and the Y (image grayscale value) IO (chromaticity information) conversion method. The method for binarizing the grayscale image may include a threshold method or an Otsu (big law) method. The method for connecting domain analysis of the binary image may include a Two-Pass (two-side scanning) algorithm or a Seed-Filling (seed filling) algorithm.

[0090] In another optional embodiment, after obtaining the binary image, the computer device can also obtain the image edge by edge detection, and then obtain the region of interest image by edge fitting and screening in combination with prior knowledge. The edge detection method may include a cross differential algorithm (Roberts) or edge detection of a first-order differential operator (Prewitt).

[0091] Step 410: Input multiple region of interest images into a first detection model to obtain initial structured information of the video to be processed.

[0092] After obtaining the region of interest image corresponding to each video frame image, the computer device inputs the multiple region of interest images into the first detection model, detects them through the first detection model, and obtains the initial structured information of the video to be processed.

[0093] In this embodiment, the first detection model is trained using the image samples of the region of interest of each video frame image corresponding to the surgical video sample and the structured information samples corresponding to the surgical video sample. This embodiment does not limit the specific process of training the first detection model, as long as its function can be achieved.

[0094] In this embodiment, the region of interest image of each video frame image is first determined, and each region of interest image is detected to obtain the initial structured information of the video to be processed. This can reduce the interference of information existing in other regions of each video frame image, thereby improving the accuracy of the determined initial structured information of the video to be processed.

[0095] In one embodiment, the initial structured information includes image-level information and time-series-level information, and the corresponding first detection model includes an image-level detection model and a time-series-level monitoring model. When acquiring image-level information, the image-level detection model in the first detection model is used; when acquiring time-series-level information, the time-series-level detection model in the first detection model is used. Image-level information refers to information acquired from a single video frame image of a video to be processed; time-series-level information refers to information acquired from multiple consecutive video frame images of a video to be processed.

[0096] In an optional embodiment, the image level information and the timing level information may be shown in the following table:

[0097]

[0098] In this case, if Figure 5 As shown, a method for implementing the method involves inputting multiple video frame images into a first detection model to obtain initial structured information of the video to be processed, and the steps of the method include:

[0099] Step 500: Input each video frame image into an image-level detection model to obtain image-level information corresponding to each video frame image.

[0100] The image-level detection model is trained using each video frame image sample corresponding to the surgical video sample, and the image-level information sample corresponding to each video frame image sample.

[0101] The computer device inputs each video frame image into the image-level detection model in sequence, detects it through the image-level detection model, and obtains the image-level information corresponding to each video frame image.

[0102] In an optional embodiment, the network architecture of the image-level detection model can be a modified Resnet network. By modifying the loss function, activation function and detection head in the Resnet network, the accuracy of the model can be improved. The network architecture of the image-level detection model can also be a CNN-RNN network, which uses a CNN network to extract semantic information from video frame images and uses an RNN network to establish the association of semantic information corresponding to each video frame image. The network architecture of the image-level detection model can also be ADDS (dual-modal classifier) ​​or a CRSA network that introduces residual attention.

[0103] Step 510: divide the multiple video frame images into multiple video frame sets according to the time sequence, and input each video frame set into the time sequence level detection model in sequence to obtain the time sequence level information corresponding to each video frame set.

[0104] The temporal level detection model is trained using the video frame set samples corresponding to the surgical video samples and the temporal level information samples corresponding to each video frame set sample.

[0105] The computer device divides the multiple video frame images into multiple video frame sets according to the time sequence. The video frame images included in each video frame set can be preset by the user, and this embodiment does not limit this.

[0106] After obtaining multiple video frame sets, the computer device sequentially inputs each video frame set into the temporal level detection model, detects them through the temporal level detection model, and obtains the temporal level information corresponding to each video frame image.

[0107] In an optional embodiment, the network architecture of the time-level detection model can be 3D CNN, CNN+RNN, RNN; the specific network model can be any one of 3D ResNet, VGG-50+LSTM, GAN+LSTM, LSTM, Deep RL.

[0108] In an optional embodiment, before inputting each video frame image into the first detection model (image-level detection model and time-series-level detection model), for each video frame image, the video frame image may be preprocessed to remove interference information in the video frame image and enhance the video frame image. This facilitates the image-level detection model to extract features of each video frame image, thereby obtaining more accurate image-level information.

[0109] In another optional embodiment, when the region of interest image is input into the first detection model (image-level detection model and time-series-level detection model), the region of interest image may be enhanced for each region of interest image before the region of interest image is input into the first detection model. Specifically, the computer device may use methods such as gamma correction to improve the contrast of the region of interest image and enhance the details of the region of interest image. This makes it easier for the image-level detection model to extract features of each video frame image, thereby obtaining more accurate image-level information.

[0110] In this embodiment, the image-level detection model and the time-level detection model in the first detection model are used to determine the image-level information and the time-level information corresponding to multiple video frame images, so that more accurate and rich initial structured information can be obtained, and then more accurate target structured information can be obtained.

[0111] In one embodiment, Figure 6As shown, it involves verifying the initial structured information according to the predefined structured information to obtain an implementation method of the target structured information of the video to be processed, and the steps of the implementation method include:

[0112] Step 600: searching for matching structured information corresponding to the initial structured information from predefined structured information; the similarity between the matching structured information and the initial structured information is greater than a preset similarity threshold.

[0113] The preset similarity threshold may be pre-stored in the computer device by the user. After obtaining the initial structured information, the computer device calculates the similarity between the initial structured information and each structured information stored in the predefined structured information, and compares each similarity with the preset similarity threshold; the structured information with a similarity greater than the preset similarity threshold is determined as the matching structured information corresponding to the initial structured information. This embodiment does not limit the specific method of calculating the similarity between the initial structured information and the predefined structured information, as long as its function can be achieved.

[0114] Step 610: Verify the initial structured information according to the matching structured information to obtain target structured information.

[0115] After determining the matching structured information, the computer device uses the matching structured information to verify the initial structured information, and the target structured information can be obtained. The matching structured information includes a complete and rich label set information corresponding to the video to be processed. The initial structured information may be modified by matching the structured information to obtain the target structured information.

[0116] In this embodiment, the matching structured information corresponding to the initial structured information is determined by calculating the similarity between the initial structured information and each structured information in the predefined structured information, and the initial structured information is modified using the matching structured information to obtain the target structured information. In this way, more accurate target structured information can be obtained, and the method for determining the target structured information is quick and easy to implement.

[0117] In one embodiment, Figure 7 As shown, an implementation method involves verifying the initial structured information according to the matching structured information to obtain the target structured information, and the steps of the implementation method include:

[0118] Step 700: Compare the initial structured information with the matching structured information to obtain a comparison result; the comparison result includes information indicating that abnormal label information exists in the initial structured information.

[0119] When the computer device uses the matching structured information to verify the initial structured information, it first compares the initial structured information with the matching structured information, that is, calculates the similarity between the initial structured information and the matching structured information to obtain a comparison result. The comparison result may include an indication that there is abnormal label information in the initial structured information, that is, the initial structured information is different from the matching structured information, or the initial structured information includes label information that does not exist in the matching structured information. The comparison result also includes the absence of abnormal label information in the initial structured information, that is, the initial structured information is the same as the matching structured information, or the label set information included in the initial structured information is less than the matching structured information.

[0120] Step 710: Verify the initial structured information according to the comparison result to obtain target structured information.

[0121] After determining the comparison result between the initial structured information and the matching structured information, the computer device verifies the initial structured information according to the comparison result to obtain the target structured information.

[0122] Specifically, when the comparison result includes the information indicating that there is abnormal label information in the initial structured information, which means that the initial structured information is different from the matching structured information, the information in the initial structured information that is different from the matching structured information is replaced with the information in the matching structured information. When the comparison result includes the information indicating that there is abnormal label information in the initial structured information, which means that the initial structured information includes label information that does not exist in the matching structured information, the redundant label information in the initial structured information is deleted.

[0123] In this embodiment, the initial structured information is verified by comparing the initial structured information with the matching structured information, so that more accurate target structured information can be obtained. In addition, the method for verifying the initial structured information in this embodiment is fast and easy to implement, which can improve the efficiency of determining the target structured information.

[0124] In an optional embodiment, it is assumed that the initial structured information includes a surgical procedure and a category of instruments corresponding to the surgical procedure. The computer device searches for matching structured information corresponding to the initial structured information from the predefined structured information; the matching structured information includes historical surgical procedures and a category of instruments corresponding to the historical surgical procedures. By comparing the category of instruments corresponding to the surgical procedure in the initial structured information with the category of instruments corresponding to the historical surgical procedures, the category of instruments corresponding to the surgical procedure in the initial structured information is verified according to the comparison result. If the comparison result indicates that there are more instruments in the instrument category in the initial structured information than in the instrument category corresponding to the historical surgical procedure, the more instruments are deleted; if the comparison result indicates that there are abnormal instruments in the instrument category in the initial structured information (that is, the instrument category in the structured information is different from the instrument in the instrument category in the matching structured information), the abnormal instruments are deleted.

[0125] In one embodiment, Figure 8 As shown, another implementation method involves verifying the initial structured information according to the predefined structured information to obtain the target structured information, and the steps of the implementation method include:

[0126] Step 800: Determine associated structured information corresponding to the initial structured information in the predefined structured information.

[0127] When the computer device verifies the initial structured information according to the structured information in the predefined structured information, it determines the associated structured information in the predefined structured information corresponding to the initial structured information. The associated structured information may be structured information in the predefined structured information that has a higher degree of association with the initial structured information.

[0128] In an optional embodiment, the computer device may determine, in the predefined structured information, structured information whose similarity with the initial structured information reaches a preset similarity threshold as the associated structured information.

[0129] Step 810: Set a ratio mark for other structured information except the associated structured information in the predefined structured information; the ratio mark is used to represent the probability of other structured information corresponding.

[0130] After determining the associated structured information, the computer device determines other information in the predefined structured information except the associated structured information as other structured information, and sets a ratio mark for the other structured information that can characterize its probability (the ratio of the other structured information in determining the target structured information). The ratio mark can be pre-set by the user for the other structured information. This embodiment does not limit the specific form of the ratio mark, as long as its function can be achieved.

[0131] Step 820: Verify the initial structured information according to the ratio identifier, the associated structured information and other structured information to obtain the target structured information.

[0132] After obtaining the scale identifier, the computer device verifies the initial structured information according to the scale identifier, the associated structured information and other structured information, and can obtain the target structured information.

[0133] In an optional embodiment, the computer device calculates a first probability that the initial structured information is associated structured information, and a second probability that the initial structured information is other structured information; calculates the product between the second probability and the ratio mark, compares the ratio of the product to the first probability with a preset probability threshold, and determines the associated structured information and other structured information greater than the preset probability threshold as the target structured information. The probability value corresponding to the ratio mark may be in the range of (probability value 0-probability value 0.5.

[0134] In an optional embodiment, the computer device may also set a ratio flag for the associated structured information, where the ratio flag is 100% or a probability value of 1.

[0135] In this embodiment, in the process of determining the target structured information, not only the associated structured information but also other structured information is considered, which can avoid missing structured information and thus improve the accuracy of the determined target structured information.

[0136] In an optional embodiment, it is assumed that the initial structured information includes a surgical procedure and a category of equipment corresponding to the surgical procedure. The computer device searches the database for associated structured information associated with the initial structured information; the associated structured information includes historical surgical procedures and a category of equipment corresponding to the surgical procedure. A ratio identifier is set for the category of equipment corresponding to the surgical procedure in other structured information other than the associated structured information in the predefined structured information. In this embodiment, the ratio identifier is a probability value. The computer device can set the ratio identifier for the equipment category in other structured information to 0, and the computer device directly verifies the initial structured information according to the associated structured information; the computer device can also set the ratio identifier for the equipment category in other structured information to 0.4, and the computer device needs to verify the initial structured information according to the ratio identifier, the associated structured information and other structured information.

[0137] In one embodiment, Fig. 9 As shown, a method for implementing the method involves inputting multiple video frame images into a first detection model to obtain initial structured information of the video to be processed, and the steps of the method include:

[0138] Step 900: Train the first detection model according to the original video and the corresponding structured information to obtain the second detection model.

[0139] The original video may be a video corresponding to the predefined structured information. The computer device obtains the original video corresponding to the predefined structured information and the structured information corresponding to the original video from the database. The first detection model is trained using the original video and the corresponding structured information to obtain the second detection model. This embodiment does not limit the specific training process as long as its function can be achieved.

[0140] In an optional embodiment, during the process of training the second detection model, a dynamic weight can be set for the structured information corresponding to the original video based on the structured information, and the dynamic weight will be dynamically adjusted based on the input information and structured information of the second detection model.

[0141] In an optional embodiment, the computer device may use the original video and the corresponding structured information to train an initial neural network model to obtain a second detection model.

[0142] Step 910: Input multiple video frame images into the second detection model to obtain initial structured information of the video to be processed.

[0143] After the computer device has trained the second detection model, it can input multiple video frame images into the second detection model, or it can input multiple video frame images into the second detection model in sequence. The input video frame images are detected by the second detection model to obtain the initial structured information of the video to be processed.

[0144] In this embodiment, the predefined structured information used in the training of the second detection model is used so that the second detection model obtained in this way can more accurately determine the initial structured information of the video to be processed, so that the target structured information after subsequent verification of the initial structured information is more accurate.

[0145] In one embodiment, after verifying the initial structured information according to the structured information, the video data processing method further includes:

[0146] The target structured information is corrected by using a time-series correction method to obtain the corrected target structured information.

[0147] After obtaining the target structured information, the computer device can also modify it for the second time, that is, use the time-series correction method to correct the target structured information, and obtain the corrected target structured information. The time-series correction method refers to correcting the obtained target structured information in time, that is, determining whether the obtained target structured information has abnormal label information in the time sequence corresponding to the video to be processed (from the beginning to the end of the video to be processed); if so, correct it.

[0148] In this embodiment, the obtained target structured information is corrected again by using a time-series correction method, so that more accurate corrected target structured information can be obtained, thereby improving the reliability of the video data processing method.

[0149] In an optional embodiment, assuming that the acquired target structured information includes information that the stage of 1h10min to 1h30min is the clip stage; the accurate information should be that the stage of 1h2min to 1h35min is the clip stage. The start and end boundaries of the stage in the information can be corrected by using a time-series correction method.

[0150] In one embodiment, Fig.10 As shown, two implementation methods of correcting the target structured information by using a time-series correction method to obtain the corrected target structured information are:

[0151] Step 101: Perform temporal smoothing processing on target structured information corresponding to a plurality of video frame images to obtain corrected target structured information.

[0152] The first time-series correction method is implemented by correcting the target structured information in the following way: the computer device determines the target structured information corresponding to the multiple video frame images arranged in time sequence in the target structured information, and performs time-series smoothing processing on the target structured information corresponding to the multiple video frame images. In other words, the computer device calculates the average value between the target structured information in a certain time sequence segment to achieve the correction of the target structured information. The specific method of time-series smoothing processing can include any one of moving average, threshold method, exponential smoothing and life card model.

[0153] In an optional embodiment, the target structured information includes image quality information of each video frame image, and the computer device determines whether the image quality is the same in a certain time sequence. For example, the image quality of the first video frame image in the time sequence is blurred, the image quality of the second video frame image is blurred; the image quality of the third video frame image is not blurred, and the image quality of the subsequent video frames is blurred; then the image quality of the third video frame image needs to be smoothed, that is, the image quality of the third video frame image is modified to blurred.

[0154] Step 102: Correct the target structured information according to the correlation between the target structured information corresponding to each video frame image to obtain corrected target structured information.

[0155] The second time-series correction method is implemented by obtaining the correlation between the target structured information at a certain time sequence. The correlation between the structured information at a certain time sequence can be a mutually dependent relationship, that is, the probability of the target structured information appearing at the same time at a certain time sequence is high; the correlation between the target structured information at a certain time sequence can also be a mutually exclusive relationship, that is, the probability of the target structured information appearing at the same time at a certain time sequence is very low.

[0156] If the computer device determines that the correlation between the target structured information corresponding to each video frame image is a mutually dependent relationship, if only one of them exists in the target structured information, the structured information with the mutual dependence relationship can be added to the target structured information to obtain the corrected target structured information. If the computer device determines that the correlation between the target structured information corresponding to each video frame image is a mutually exclusive relationship, if two structured information with mutually exclusive relationship exist in the target structured information, one of the two structured information is removed; specifically, the one with a lower probability of the two structured information can be removed.

[0157] In an optional embodiment, the computer device pre-stores a mutual dependency matrix and a mutual ordering matrix between the structured information. The target structured information is corrected according to the matrix to obtain the corrected target structured information.

[0158] In an optional embodiment, assuming that the surgical procedure included in the acquired target structured information is cholecystectomy, and the corresponding surgical instruments include surgical forceps and a recovery bag, then the probability of being in the gallbladder recovery stage at this time is extremely high; then the timing boundary in the target structured information can be corrected based on the detection of the surgical instruments in the target structured information. In other words, the surgical forceps and the recovery bag are interdependent. If the target structured information in a certain time segment includes surgical forceps and a recovery bag, it means that the time segment represents the gallbladder recovery stage, that is, the timing including surgical forceps and recovery bags in the target structured information is updated to the gallbladder recovery stage.

[0159] In another optional embodiment, it is assumed that the instrument categories in the acquired target structured information cannot contain both ultrasonic scalpel and electrocoagulation rod, that is, ultrasonic scalpel and electrocoagulation rod are related and exclusive structured information. If both ultrasonic scalpel and electrocoagulation rod exist in the target structured information, then the target structured information is removed from the target structured information, and the corrected target structured information is obtained.

[0160] In this embodiment, two specific methods for correcting target structured information using a time-series correction method are provided, and the user can select one or two of them to correct the target structured information according to actual needs, which can improve the flexibility and practicality of video data processing.

[0161] In one embodiment, after correcting the initial structured information according to the structured information, the video data processing method further includes:

[0162] A correspondence between the target structured information and the video to be processed is established, and the correspondence is stored.

[0163] After obtaining the target structured information, the computer device establishes a corresponding relationship between the target structured information and the video to be processed, and stores the target structured information, the video to be processed and the corresponding relationship in a database.

[0164] If the computer device uses a time-level correction method to perform time-series smoothing on the target structured information to obtain corrected target structured information, the corrected target structured information will be used to replace the corresponding target structured information stored in the database.

[0165] In this embodiment, by storing the correspondence between the target structured information and the video to be processed, it is convenient to quickly obtain the target structured information corresponding to the video to be processed later, and analyze the target structured information to realize more upper-level applications.

[0166] In an optional embodiment, the specific upper-layer application may include statistics on the instrument category, operation duration and other information included in the target structured information to facilitate medical staff to more accurately select the correct instrument for the surgical procedure; at the same time, the doctor's skill level can also be analyzed based on the statistical information such as the instrument category and operation duration. The upper-layer application may also include repairing low-quality structured information through relevant algorithms according to the image quality included in the target structured information (applying a deblurring algorithm to a blurred image, and applying a smoke algorithm to remove smoke when the image quality is smoke). The upper-layer application may also include scoring and evaluating the quality of the surgery by analyzing the time consumed in each surgical stage in the target structured information and the existing surgical abnormalities. The upper-layer application can also allow users to search for corresponding videos and other target structured information from the database based on the input video-related information. For example, the user can search for multiple corresponding surgical videos in the database by entering patient information, doctor information and surgical procedure information: [50-60 years old, female, chief physician, cholecystectomy, scissors].

[0167] See also Fig.11 An embodiment of the present application provides a video data processing method, the method comprising the steps of:

[0168] Step 110: Obtain the video to be processed;

[0169] Step 111: extract multiple video frame images from the video to be processed according to a preset extraction rule;

[0170] Step 112, identifying the region of interest in each video frame image to obtain the region of interest image corresponding to each video frame image;

[0171] Step 113: input multiple region of interest images into an image-level detection model to obtain image-level information corresponding to each region of interest image;

[0172] Step 114, dividing the plurality of region of interest images into a plurality of region of interest sets according to the time sequence, and sequentially inputting the region of interest image sets into the time sequence level detection model to obtain the time sequence level information corresponding to each region of interest image;

[0173] Step 115: searching for matching structured information that matches the initial structured information from predefined structured information; the initial structured information includes image-level information and time-series-level information, and the similarity between the matching structured information and the initial structured information is greater than a preset similarity threshold;

[0174] Step 116: compare the initial structured information with the matching structured information to obtain a comparison result; the comparison result includes information indicating that abnormal label information exists in the initial structured information;

[0175] Step 117: verify the initial structured information according to the comparison result to obtain target structured information;

[0176] Step 118: Correct the target structured information using a time-series correction method to obtain corrected target structured information; the time-series correction method includes performing time-series smoothing processing on the target structured information corresponding to the multiple video frame images; and correcting the target structured information according to the correlation between the target structured information corresponding to each video frame image;

[0177] Step 119: establish a corresponding relationship between the corrected target structured information and the video to be processed, and store the corresponding relationship.

[0178] In an optional embodiment, an example of target structured information is shown in the following table:

[0179]

[0180] The information that users can obtain through the above target structured information is: the leading doctor of the operation is a chief physician, the name of the hospital is Peking Union Medical College, which is a tertiary hospital. The patient is 45 years old, male, with a BMI index of 29.1 and an obese body. In terms of video quality, the first frame of the video is blurry and there is smoke, the second and third frames are blurry; there is smoke in the fourth frame, etc. In terms of instrument category, there are scissors in the first and second frames of the video, and there are separation forceps in the third frame. In terms of anatomical structure, the gallbladder structure is included in the first and second frames of the video. In terms of surgical stage, the first three frames of the video belong to the clipping stage. In terms of abnormal hand event category, there is abnormal information of bleeding and unclean cutting edge in the third stitch of the video.

[0181] In an optional embodiment, the schematic diagram of the target structured information is as follows: Fig.12 shown.

[0182] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0183] Based on the same inventive concept, the embodiment of the present application also provides a video data processing device for implementing the video data processing method involved above. The implementation solution provided by the device to solve the problem is similar to the implementation solution recorded in the above method, so the specific limitations in the one or more video data processing device embodiments provided below can refer to the limitations of the video data processing method above, and will not be repeated here.

[0184] In one embodiment, Fig.13 As shown, a video data processing device 10 is provided, comprising: an acquisition module 11, a determination module 12 and a verification module 13, wherein:

[0185] The acquisition module 11 is used to acquire the video to be processed.

[0186] The determination module 12 is used to determine the initial structured information of the video to be processed according to the video to be processed; the initial structured information includes the tag set information to be processed associated with the video to be processed.

[0187] The verification module 13 is used to verify the initial structured information according to the predefined structured information to obtain the target structured information of the video to be processed; wherein the predefined structured information includes the label set information associated with the surgical video stored in the database.

[0188] The video data processing device provided in the above embodiment has an implementation principle and technical effects similar to those of the above method embodiment, which will not be described in detail here.

[0189] Each module in the above-mentioned video data processing device can be implemented in whole or in part by software, hardware or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module above.

[0190] In one embodiment, a computer device is provided, and when the computer program is executed by a processor, a video data processing method is implemented.

[0191] Those skilled in the art will understand that Figure 1 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0192] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0193] Get the video to be processed;

[0194] According to the video to be processed, initial structured information of the video to be processed is determined; the initial structured information includes information of a tag set to be processed associated with the video to be processed;

[0195] The initial structured information is verified according to the predefined structured information to obtain the target structured information of the video to be processed; wherein the predefined structured information includes the label set information associated with the surgical video stored in the database.

[0196] The above embodiment provides a computer device, whose implementation principle and technical effect are similar to those of the above method embodiment, and will not be repeated here.

[0197] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0198] Get the video to be processed;

[0199] According to the video to be processed, initial structured information of the video to be processed is determined; the initial structured information includes information of a tag set to be processed associated with the video to be processed;

[0200] The initial structured information is verified according to the predefined structured information to obtain the target structured information of the video to be processed; wherein the predefined structured information includes the label set information associated with the surgical video stored in the database.

[0201] The above embodiment provides a computer-readable storage medium, whose implementation principle and technical effect are similar to those of the above method embodiment, and will not be repeated here.

[0202] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:

[0203] Get the video to be processed;

[0204] According to the video to be processed, initial structured information of the video to be processed is determined; the initial structured information includes information of a tag set to be processed associated with the video to be processed;

[0205] The initial structured information is verified according to the predefined structured information to obtain the target structured information of the video to be processed; wherein the predefined structured information includes the label set information associated with the surgical video stored in the database.

[0206] The above embodiment provides a computer program product, whose implementation principle and technical effect are similar to those of the above method embodiment, and will not be repeated here.

[0207] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0208] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0209] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A video data processing method, characterized in that: The method comprises: Get the video to be processed; Determining initial structured information of the video to be processed according to the video to be processed; the initial structured information includes information of a tag set to be processed associated with the video to be processed; The initial structured information is verified according to predefined structured information to obtain target structured information of the video to be processed; wherein the predefined structured information includes label set information associated with the surgical video stored in a database.

2. The method according to claim 1, characterized in that The step of determining the initial structured information of the video to be processed according to the video to be processed includes: Extracting a plurality of video frame images from the video to be processed according to a preset extraction rule; Input the plurality of video frame images into a first detection model to obtain the initial structural information of the video to be processed.

3. The method according to claim 2, characterized in that The method further comprises: Identify the region of interest in each of the video frame images to obtain a region of interest image corresponding to each of the video frame images; The plurality of region of interest images are input into a first detection model to obtain the initial structured information of the video to be processed.

4. The method according to claim 2, characterized in that: The initial structured information includes image-level information and time-series-level information, the first detection model includes an image-level detection model and a time-series-level detection model, and the inputting of the plurality of video frame images into the first detection model to obtain the initial structured information of the video to be processed includes: Inputting each of the video frame images into the image-level detection model to obtain image-level information corresponding to each of the video frame images; The plurality of video frame images are divided into a plurality of video frame sets according to time sequence, and each of the video frame sets is sequentially input into the time sequence level detection model to obtain time sequence level information corresponding to each of the video frame sets.

5. The method according to any one of claims 1 to 4, characterized in that: The step of verifying the initial structured information according to the predefined structured information to obtain target structured information of the video to be processed includes: Searching for matching structured information corresponding to the initial structured information from the predefined structured information; the similarity between the matching structured information and the initial structured information is greater than a preset similarity threshold; The initial structured information is verified according to the matching structured information to obtain the target structured information.

6. The method according to claim 5, characterized in that The verifying the initial structured information according to the matching structured information to obtain the target structured information includes: Comparing the initial structured information with the matching structured information to obtain a comparison result; the comparison result includes information indicating that abnormal label information exists in the initial structured information; The initial structured information is verified according to the comparison result to obtain the target structured information.

7. The method according to any one of claims 1 to 4, characterized in that: The initial structured information is verified according to the predefined structured information to obtain the target structured information of the video to be processed, including: Determining associated structured information in the predefined structured information that corresponds to the initial structured information; Setting a ratio mark for other structured information in the predefined structured information except the associated structured information; the ratio mark is used to represent the probability corresponding to the other structured information; The initial structured information is detected according to the ratio identifier, the associated structured information and the other structured information to obtain the target structured information.

8. The method according to any one of claims 2 to 4, characterized in that: The step of inputting the plurality of video frame images into a first detection model to obtain the initial structured information of the video to be processed includes: Training the first detection model according to the original video and the corresponding structured information to obtain a second detection model; Input the plurality of video frame images into the second detection model to obtain the initial structural information of the video to be processed.

9. The method according to any one of claims 1 to 4, characterized in that: The method further comprises: The target structured information is corrected by using a time-series correction method to obtain corrected target structured information.

10. The method according to claim 9, characterized in that The method of correcting the target structured information by using a time-series correction method to obtain the corrected target structured information includes: Performing temporal smoothing processing on the target structured information corresponding to the plurality of video frame images to obtain the corrected target structured information; and / or; According to the correlation between the target structured information corresponding to each of the video frame images, the target structured information is corrected to obtain the corrected target structured information.

11. The method according to any one of claims 1 to 4, characterized in that: The method further comprises: A corresponding relationship between the target structured information and the video to be processed is established, and the corresponding relationship is stored.

12. A video data processing device, characterized in that: The device comprises: An acquisition module is used to acquire the video to be processed; A determination module, configured to determine initial structured information of the video to be processed according to the video to be processed; the initial structured information includes information of a tag set to be processed associated with the video to be processed; A verification module is used to verify the initial structured information according to predefined structured information to obtain target structured information of the video to be processed; wherein the predefined structured information includes label set information associated with the surgical video stored in a database.

13. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 11 are implemented.

14. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 11 are implemented.

15. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 11 are implemented.

Images