Data collaboration verification method and system based on intelligent AI

Abstract

This invention relates to a data collaborative verification method based on intelligent AI, belonging to the field of electronic digital data processing, and more specifically to the field of machine learning. The method includes: acquiring a current data segment to be verified by AI, wherein the current data segment is represented by a customized data structure at the underlying data level; and using an AI data collaborative verification model to collaboratively verify the existence of the target person in the current data segment and the frame number of that person based on the current data segment and the corresponding digital file of the target person's baseline body shape pattern. This invention addresses the technical problem of the complex and cumbersome process of verification using non-underlying data in existing technologies. By representing massive amounts of images in a customized data structure at the underlying data level, supplemented by other basic data, and then utilizing an artificial intelligence model obtained based on machine learning, the invention completes the synchronous intelligent verification of the existence of the target person in massive amounts of images and the specific frame in which they reside, thereby solving the aforementioned technical problems.

Description

Technical Field

[0001] The present invention relates to the processing of electronic digital data, more specifically to the field of machine learning, and particularly to a data collaborative verification method and system based on intelligent AI. Background Technology

[0002] Machine learning is a subfield of digital data processing. It's a technique that uses algorithms and models to enable computers to automatically learn from data and make predictions or analyses, falling under the category of artificial intelligence. The core goal of machine learning is to allow computers to analyze large amounts of data, identify patterns and rules, and thus build models that adapt to new data, without explicit programming instructions.

[0003] Machine learning has adaptive, automated, and generalization capabilities, including different types such as supervised learning, unsupervised learning, and reinforcement learning. It is widely used in application scenarios such as image recognition, natural language processing, recommendation systems, and autonomous driving. For example, it is hoped that artificial intelligence models obtained based on machine learning can be used to intelligently verify the existence of various features in massive images represented in data form and the specific location of the frames in which various features are located.

[0004] For example, Chinese invention patent publication CN113744418A proposes a ticket verification system based on an edge-cloud integrated environment. The system includes a database, a collaboration module, a data acquisition module, a verification module, a processing module, an association module, and a processor. The collaboration module coordinates the verification data from the verification module with external data to determine the association between them. The data acquisition module collects user data to verify ticket data. The verification module checks whether the user's ticket information is associated with their current itinerary to verify the user's itinerary. The processing module processes the collected data and feeds it back to the association module for further association. This invention uses the association module to associate the ticket with the user's identity and their itinerary data. After the user's data is associated, efficient verification of the user's data is achieved.

[0005] For example, Chinese invention patent publication CN112668601A proposes an image recognition system and method based on machine learning. The system includes: a location information acquisition unit for acquiring location information representing the current location of the object to be recognized; an image information acquisition unit for acquiring image information surrounding the object to be recognized; an attribute information acquisition unit for acquiring attribute information of the object to be recognized based on the location information; an object category determination unit for determining the object category, which is one or more categories that are objects for image recognition processing of the image information; and an image recognition unit for using the object category determined by the object category determination unit as the object and performing image recognition processing on the object in the image information. The system has the advantages of accurate recognition, high precision, and high efficiency.

[0006] However, the existing data collaborative verification solutions all focus on using application-level data and verification mechanisms, without utilizing underlying data and verification mechanisms. This results in a complex and cumbersome verification process. Therefore, a data collaborative verification solution based on various underlying data is needed to intelligently verify the existence and location of features in application scenarios. For example, a solution is needed that can intelligently verify the existence of various features in massive images represented in the form of underlying data, using artificial intelligence models obtained through machine learning, and avoid getting bogged down in complex and cumbersome application-level data processing. Summary of the Invention

[0007] To address the technical problems in existing technologies, this invention provides a data collaborative verification method and system based on intelligent AI. This method, based on a customized data structure representing a massive amount of video data, is supplemented by various other basic data and utilizes an AI model obtained through machine learning to simultaneously and intelligently verify the existence of each designated person target within the massive video data and the specific frame in which the target is located. Crucially, it sets up AI models with different customized structures for different formats of personnel baseline body shape patterns. This ensures the effectiveness and stability of the intelligent verification results while enabling personnel detection and location across file formats using non-image recognition modes, avoiding the complex and tedious process of massive image recognition.

[0008] According to a first aspect of the present invention, a data collaborative verification method based on intelligent AI is provided, the method comprising:

[0009] The current data segment to be verified by AI is obtained. The current data segment is obtained by concatenating the data of multiple digital files corresponding to multiple frames with the same resolution and the same contrast. The multiple digital files all adopt the first image file format and the multiple frames are assigned their respective serial numbers according to the order of concatenation.

[0010] The data of the digital file corresponding to the baseline body shape pattern of the target person is obtained, and the digital file corresponding to the baseline body shape pattern of the target person adopts the second image file format;

[0011] Obtain the monitoring scene association information corresponding to the current data segment to be verified by AI;

[0012] Perform each learning action on the multilayer feedforward neural network to obtain the multilayer feedforward neural network after each learning action is completed, and use it as an AI data collaborative verification model.

[0013] The AI ​​data collaborative verification model is adopted to verify the current data segment to be verified by AI, its corresponding resolution, contrast, and the data of each monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment. The model is used to collaboratively verify the number of frames in the current data segment in which the target person appears, as well as the multiple serial numbers corresponding to the multiple frames in which the target person appears.

[0014] Among them, the number of times the learning action is performed follows a positive correlation with the total number of pixels in the baseline body shape pattern.

[0015] Among them, the number of frames in a multi-frame image is greater than or equal to a preset frame number threshold.

[0016] According to a second aspect of the present invention, a data collaborative verification system based on intelligent AI is provided, the system comprising a memory and a plurality of processors, the memory storing a computer program configured to be executed by the plurality of processors to complete the following steps:

[0017] The current data segment to be verified by AI is obtained. The current data segment is obtained by concatenating the data of multiple digital files corresponding to multiple frames with the same resolution and the same contrast. The multiple digital files all adopt the first image file format and the multiple frames are assigned their respective serial numbers according to the order of concatenation.

[0018] The data of the digital file corresponding to the baseline body shape pattern of the target person is obtained, and the digital file corresponding to the baseline body shape pattern of the target person adopts the second image file format;

[0019] Obtain the monitoring scene association information corresponding to the current data segment to be verified by AI;

[0020] Perform each learning action on the multilayer feedforward neural network to obtain the multilayer feedforward neural network after each learning action is completed, and use it as an AI data collaborative verification model.

[0021] The AI ​​data collaborative verification model is adopted to verify the current data segment to be verified by AI, its corresponding resolution, contrast, and the data of each monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment. The model is used to collaboratively verify the number of frames in the current data segment in which the target person appears, as well as the multiple serial numbers corresponding to the multiple frames in which the target person appears.

[0022] Among them, the number of times the learning action is performed follows a positive correlation with the total number of pixels in the baseline body shape pattern.

[0023] Among them, the number of frames in a multi-frame image is greater than or equal to a preset frame number threshold.

[0024] According to a third aspect of the present invention, a data collaborative verification system based on intelligent AI is provided, the system comprising:

[0025] The first capturing device is used to acquire the current data segment to be verified by AI. The current data segment is obtained by connecting the data of multiple digital files corresponding to multiple frames with the same resolution and the same contrast. The multiple digital files all adopt the first image file format and the multiple frames are assigned their respective serial numbers according to the order of connection.

[0026] The second capturing device is used to acquire data of a digital file corresponding to the baseline body shape pattern of the target person. The digital file corresponding to the baseline body shape pattern of the target person adopts a second image file format.

[0027] The third capture device is used to acquire the monitoring scene association information corresponding to the current data segment to be verified by AI.

[0028] Machine learning devices are used to perform various learning actions on a multi-layer feedforward neural network to obtain a multi-layer feedforward neural network after each learning action is performed, and to serve as an AI data collaborative verification model.

[0029] The collaborative verification device is connected to the first capture device, the second capture device, and the third capture device respectively. It is used to use the AI ​​data collaborative verification model to verify the current data segment to be verified by AI, its corresponding resolution, contrast, and the data of each monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment. It also verifies the number of frames in the current data segment in which the target person appears and the multiple serial numbers corresponding to the multiple frames in which the target person appears.

[0030] Among them, the number of times the learning action is performed follows a positive correlation with the total number of pixels in the baseline body shape pattern.

[0031] Among them, the number of frames in a multi-frame image is greater than or equal to a preset frame number threshold.

[0032] Compared with the prior art, the present invention has at least the following five key inventive points:

[0033] The first invention point: replacing the complex and cumbersome frame-by-frame image recognition mechanism, for massive image frames with a wide time range, it converts them into current data segments with customized data structures to be verified by AI data. The intelligent AI-based data collaborative verification mechanism uses the current data segment and various auxiliary data to determine whether different people exist in the massive image frames and, if they exist, to determine the specific location of multiple frames. Different people use their own reference body shape patterns as the data collaborative verification objects. The resolution and contrast of their respective reference body shape patterns are the same. The digital file format of the massive image frames and the digital file format of the reference body shape patterns can be different, thereby realizing personnel detection and personnel location in a non-image recognition mode that crosses file formats, avoiding the complex and cumbersome massive image recognition.

[0034] The second invention point: AI data collaborative verification models with different customized structures are designed for the baseline body shape patterns of different image standards. These models are used to complete the intelligent verification of the presence and specific location of people in a large number of image frames using a data collaborative verification mechanism under intelligent AI. The AI ​​data collaborative verification model is a multi-layer feedforward neural network after each learning action. The multi-layer feedforward neural network includes a single input layer, multiple hidden layers, and a single output layer. The number of hidden layers follows the numerical trend of the contrast of the baseline body shape pattern, and the number of learning actions follows the numerical trend of the positive correlation with the total number of pixels in the baseline body shape pattern. The customized structural designs in all these places ensure the effectiveness and stability of the verification results.

[0035] The third invention point: In order to complete the intelligent verification of the presence and specific location of people in a large number of video frames by adopting a data collaborative verification mechanism under intelligent AI, a variety of different basic data are introduced to participate in the data collaborative verification. The various basic data include the current data segment to be verified by AI and its corresponding resolution, contrast and each monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the person to be identified, the image file format used in the current data segment and the two format type numbers corresponding to the image file format used in the baseline body shape pattern, and the number of all video frames in the current data segment. The full and comprehensive selection of the above-mentioned various basic data further ensures the effectiveness and stability of the verification results.

[0036] The fourth point of invention: Specifically, the current data segment to be verified by AI is obtained by concatenating the data of multiple digital files corresponding to multiple frames with the same resolution and contrast. The multiple frames are assigned their own serial numbers according to the order of concatenation. The scene monitoring camera from which the multiple frames in the current data segment to be verified by AI is taken as the current scene camera. The shooting frame rate, exposure time, aperture value and shutter speed of the current scene camera, as well as the monitoring area area corresponding to the monitoring scene where the current scene camera is located, are taken as the monitoring scene association information corresponding to each monitoring scene in the current data segment to be verified by AI. Thus, the data structure design of various basic data for intelligent verification is completed.

[0037] The fifth inventive point: In each learning action performed on the multilayer feedforward neural network, the number of frames in a known data segment showing a certain person, and the multiple sequence numbers corresponding to the multiple frames showing the person, are used as the output content of the multilayer feedforward neural network. The data segment and its corresponding resolution, contrast, and each monitoring scene association information, the data of the digital file corresponding to the reference body shape pattern of the person, the format type number corresponding to the image file format used by each frame in the data segment, the format type number corresponding to the image file format used by the reference body shape pattern of the person, and the total number of frames in the data segment are used as the input content of the multilayer feedforward neural network to complete the learning action, thereby ensuring the learning effect of each learning action of the multilayer feedforward neural network. Attached Figure Description

[0038] The embodiments of the present invention will now be described with reference to the accompanying drawings, wherein:

[0039] Figure 1 This is a schematic diagram of the working scenario of the data collaborative verification method and system based on intelligent AI according to the present invention.

[0040] Figure 2This is a flowchart illustrating the steps of a data collaborative verification method based on intelligent AI according to Embodiment 1 of the present invention.

[0041] Figure 3 The following is a flowchart illustrating the steps of a data collaborative verification method based on intelligent AI according to Embodiment 2 of the present invention.

[0042] Figure 4 The following is a flowchart illustrating the steps of a data collaborative verification method based on intelligent AI according to Embodiment 3 of the present invention.

[0043] Figure 5 The following is a flowchart illustrating the steps of a data collaborative verification method based on intelligent AI according to Embodiment 4 of the present invention.

[0044] Figure 6 This is a schematic diagram of the structure of a data collaborative verification system based on intelligent AI, as shown in Embodiment 5 of the present invention.

[0045] Figure 7 This is a schematic diagram of the structure of a data collaborative verification system based on intelligent AI, as shown in Embodiment 6 of the present invention. Detailed Implementation

[0046] like Figure 1 The diagram illustrates a working scenario of the data collaborative verification method and system based on intelligent AI according to the present invention. The present invention relates to the processing of electronic digital data, and more specifically to the field of machine learning.

[0047] The specific technical process of this invention is as follows:

[0048] Technical Process A: Design AI data collaborative verification models with different customized structures for the baseline body shape patterns of different image standards. These models are used to complete the intelligent verification of the presence of people in massive image frames and the specific location of people using a data collaborative verification mechanism under intelligent AI.

[0049] Specifically, the definition of whether two frames have the same image standard is completed by using the same resolution and the same contrast. That is, when the resolution and contrast of two frames are the same, it is determined that the two frames have the same image standard.

[0050] Customized structural design 1. The AI ​​data collaborative verification model is a multi-layer feedforward neural network after each learning action is completed. The number of learning actions completed by the multi-layer feedforward neural network follows the numerical change trend of the total number of pixels in the baseline body shape pattern in a positive correlation.

[0051] Customized structural design 2: The multi-layer feedforward neural network used includes a single input layer, multiple hidden layers and a single output layer, and the number of hidden layers follows the numerical trend of the contrast of the baseline body shape pattern.

[0052] In the customized structural design, in each learning action of the multilayer feedforward neural network, the number of frames in a known data segment showing a certain person, and the multiple sequence numbers corresponding to the multiple frames showing the person, are used as the output content of the multilayer feedforward neural network. The data segment and its corresponding resolution, contrast, and monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the person, the format type number corresponding to the image file format used by each frame in the data segment, the format type number corresponding to the image file format used by the baseline body shape pattern of the person, and the total number of frames in the data segment are used as the input content of the multilayer feedforward neural network to complete the learning action, thereby ensuring the learning effect of each learning action of the multilayer feedforward neural network.

[0053] In this way, through the above three customized structural designs, AI data collaborative verification models with different customized structures are designed for the baseline body shape patterns of different image standards, ensuring the validity and stability of the verification results;

[0054] Obviously, when it is necessary to perform existence analysis and location analysis under the underlying data collaborative verification mode for different people, if the different baseline body shape patterns corresponding to different people have the same format, then the analysis of these different people can use the same AI data collaborative verification model.

[0055] Technical Process B: In order to complete the intelligent verification of the presence and specific location of people in a large number of video frames by adopting a data collaborative verification mechanism under intelligent AI, a variety of different basic data were introduced to participate in the data collaborative verification.

[0056] Specifically, the various basic data include the current data segment to be verified by AI data and its corresponding resolution, contrast, and each set of monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the person to be identified, the image file format used in the current data segment and the two format type numbers corresponding to the image file format used in the baseline body shape pattern, and the number of all frames in the current data segment.

[0057] like Figure 1 As shown, the various basic data include the pattern data of the person to be identified, namely the digital file data corresponding to the baseline body shape pattern of the person to be identified, the current data segment to be verified by AI data, and other auxiliary data, namely the resolution, contrast and each monitoring scene association information corresponding to the current data segment, the image file format used in the current data segment and the two format type numbers corresponding to the image file format used in the baseline body shape pattern, and the number of all frames in the current data segment.

[0058] More specifically, the current data segment to be verified by AI is obtained by concatenating the data of multiple digital files corresponding to multiple frames with the same resolution and contrast. The multiple frames are assigned their own serial numbers according to the order of concatenation. The scene monitoring camera from which the multiple frames in the current data segment to be verified by AI is taken as the current scene camera. The shooting frame rate, exposure time, aperture value, shutter speed of the current scene camera, and the monitoring area area corresponding to the monitoring scene where the current scene camera is located are taken as the monitoring scene association information corresponding to each monitoring scene in the current data segment to be verified by AI. Thus, the data structure design of various basic data for intelligent verification is completed.

[0059] In this way, by fully and comprehensively selecting from the above-mentioned various basic data, the validity and stability of the verification results are further guaranteed;

[0060] Technical Process C: The AI ​​data collaborative verification model, which adopts the customized structure design of Technical Process A, intelligently verifies the presence of people and their specific locations within a large number of video frames based on multiple basic data selected in Technical Process B.

[0061] Specifically, the intelligent verification result is whether the set person exists in the massive number of screen frames, and if so, it also provides the screen number corresponding to one or more screens in which the set person exists.

[0062] Clearly, here the existence and location of any person can be determined through the intelligent verification mode of the underlying data in any sequence of images, avoiding the use of more complex and cumbersome application layer data processing;

[0063] Technical Process D: Based on the intelligent verification results of Technical Process C, redundant data segments are represented for personnel in a massive number of screen frames to help personnel or machines skip the massive number of screen frames and move to the next massive number of screen frames to search for personnel.

[0064] For example, when the number of frames of the target person in the current data segment to be verified by AI is zero, the current data segment is marked as a redundant data segment for the target person; when the number of frames of the target person in the current data segment to be verified by AI is greater than zero, the current data segment is marked as a non-redundant data segment for the target person.

[0065] Therefore, this invention can replace the complex and cumbersome frame-by-frame image recognition mechanism. For massive amounts of image frames with a wide time range, it converts them into current data segments with customized data structures to be verified by AI. The intelligent AI-based data collaborative verification mechanism uses the current data segment and various auxiliary data to determine whether different people exist in the massive amount of image frames, and if so, to locate them in multiple frames. Different people use their own baseline body shape patterns as the data collaborative verification objects. The resolution and contrast of each baseline body shape pattern are the same. The digital file format of the massive amount of image frames and the digital file format of the baseline body shape patterns can be different, thereby realizing personnel detection and location in a non-image recognition mode that transcends file formats, avoiding the complex and cumbersome massive image recognition.

[0066] The key points of this invention are: the directional design of AI data collaborative verification models with different customized structures for baseline body shape patterns of different image standards; the full and comprehensive selection of various basic data participating in AI data collaborative verification; the targeted design of each learning action of the multi-layer feedforward neural network; personnel detection and personnel positioning in non-image recognition modes across file formats; the marking of redundant data segments of set personnel based on intelligent verification results; and the skipping processing of set personnel search in the corresponding massive number of image frames.

[0067] The data collaborative verification method and system based on intelligent AI of the present invention will be specifically described below by way of embodiments. Example 1

[0068] Figure 2 This is a flowchart illustrating the steps of a data collaborative verification method based on intelligent AI according to Embodiment 1 of the present invention.

[0069] like Figure 2 As shown, the data collaborative verification method based on intelligent AI includes the following specific steps:

[0070] Step 201: Obtain the current data segment to be verified by AI. The current data segment is obtained by concatenating the data of multiple digital files corresponding to multiple frames with the same resolution and contrast. The multiple digital files all adopt the first image file format and the multiple frames are assigned their respective serial numbers according to the order of concatenation.

[0071] Specifically, image file format refers to the storage format of the underlying data of a computer image. Common image file formats include bmp, jpeg, png, tif, gif, pcx, tga, exif, fpx, svg, psd, cdr, pcd, dxf, ufo, eps, ai, raw, WMF, webp, avif, apng, etc.

[0072] More specifically, BMP format, or bitmap format, is a hardware-independent image file format that is widely used. It uses a bitmap storage format and does not use any other compression except for the selectable image depth. The image depth of BMP file format can be 1 bit, 4 bit, 8 bit, or 24 bit. When storing data in BMP file format, the image is scanned in a left-to-right, bottom-to-top order.

[0073] More specifically, the JPEG format, also known as the Joint Photo Experts Group format, is the most common image format. It is a lossy compression format that can compress images into a smaller storage space. It uses lossy compression to remove redundant image data, achieving a very high compression ratio while displaying rich and vivid images. In other words, it can obtain good image quality with minimal disk space. The JPEG file format mainly compresses high-frequency information and retains color information well. It is suitable for use on the Internet, can reduce image transmission time, can support 24-bit true color, and is also widely used for images that require continuous tones.

[0074] Step 202: Obtain the data of the digital file corresponding to the baseline body shape pattern of the target person. The digital file corresponding to the baseline body shape pattern of the target person adopts the second image file format.

[0075] Clearly, this enables personnel detection and localization in non-image recognition modes across file formats because the baseline body shape pattern of the target personnel participating in data collaborative verification and the image in the current data segment to be verified by AI can use different image file formats.

[0076] For example, the baseline body shape image of the target person can be in jpeg format, while the image in the current data segment to be verified by AI can be in bmp format;

[0077] Step 203: Obtain the monitoring scene association information corresponding to the current data segment to be verified by AI;

[0078] Specifically, the monitoring scene association information corresponding to the current data segment to be verified by AI includes the camera parameters of the camera device that performs the monitoring screen acquisition, as well as the actual scene information of the monitoring scene;

[0079] Step 204: Perform each learning action on the multilayer feedforward neural network to obtain the multilayer feedforward neural network after each learning action is performed, and use it as an AI data collaborative verification model;

[0080] Specifically, numerical simulation mode can be used to complete the testing and simulation of the model building process of performing each learning action on the multilayer feedforward neural network to obtain the multilayer feedforward neural network after performing each learning action, and to serve as the model building process of the AI ​​data collaborative verification model.

[0081] Step 205: Using the AI ​​data collaborative verification model, based on the current data segment to be verified by AI and its corresponding resolution, contrast, and the data of each monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment, collaboratively verify the number of frames in the current data segment in which the target person appears, and the multiple serial numbers corresponding to the multiple frames in which the target person appears.

[0082] Among them, the number of times the learning action is performed follows a positive correlation with the total number of pixels in the baseline body shape pattern.

[0083] For example, the total number of pixels in the baseline body shape pattern is 1280 pixels × 720 pixels, and the number of times the learning action is performed is 800; the total number of pixels in the baseline body shape pattern is 1920 pixels × 1080 pixels, and the number of times the learning action is performed is 900; the total number of pixels in the baseline body shape pattern is 3840 pixels × 2160 pixels, and the number of times the learning action is performed is 1000, and so on.

[0084] Among them, the number of frames in a multi-frame scene is greater than or equal to a preset frame number threshold;

[0085] Specifically, the preset frame rate threshold is relatively large, for example, 100,000 frames. In this way, the multi-frame images used for search personnel can be determined to be a massive number of frames.

[0086] In each learning action performed on the multilayer feedforward neural network, the number of frames in a known data segment showing a certain person and the multiple sequence numbers corresponding to the multiple frames showing the person are used as the output content of the multilayer feedforward neural network. The data segment and its corresponding resolution, contrast, and each monitoring scene association information, the data of the digital file corresponding to the reference body shape pattern of the person, the format type number corresponding to the image file format used by each frame in the data segment, the format type number corresponding to the image file format used by the reference body shape pattern of the person, and the total number of frames in the data segment are used as the input content of the multilayer feedforward neural network to complete this learning action.

[0087] Among them, obtaining the monitoring scene association information corresponding to the current data segment to be verified by AI includes: taking the scene monitoring camera from which the multiple frames in the current data segment to be verified by AI as the current scene camera, and taking the shooting frame rate, exposure time, aperture value and shutter speed of the current scene camera and the monitoring area corresponding to the monitoring scene where the current scene camera is located as the monitoring scene association information corresponding to the current data segment to be verified by AI.

[0088] Among them, the data of the digital file corresponding to the baseline body shape pattern of the target personnel is obtained. The digital file corresponding to the baseline body shape pattern of the target personnel adopts the second image file format, which includes: the resolution and contrast of each baseline body shape pattern corresponding to each person are the same.

[0089] Specifically, the definition of whether two frames have the same image standard is completed by using the same resolution and the same contrast. That is, when the resolution and contrast of two frames are the same, it is determined that the two frames have the same image standard.

[0090] The current data segment is obtained by concatenating the data of multiple digital files corresponding to multiple frames with the same resolution and contrast. The multiple digital files all adopt the first image file format, and the multiple frames are assigned their own serial numbers according to the order of concatenation. The multiple frames are assigned their own serial numbers in the current data segment according to the order of concatenation of the data of their respective multiple digital files in the current data segment.

[0091] Specifically, for multiple frames within the current data segment, different binary values, from smallest to largest, are used to represent the multiple frame sequence numbers corresponding to each frame.

[0092] And wherein, the sequence number of each of the multiple frames is obtained in the current data segment according to the order in which the data of their respective multiple digital files are connected end to end in the current data segment, including: for each frame in the multiple frames, the earlier the data of its corresponding digital file participates in the connection in the current data segment, the smaller its sequence number in the current data segment. Example 2

[0093] Figure 3 The following is a flowchart illustrating the steps of a data collaborative verification method based on intelligent AI according to Embodiment 2 of the present invention.

[0094] like Figure 3 As shown, with Figure 2Unlike the previous implementation, after using an AI data collaborative verification model to collaboratively verify the current data segment to be verified by AI, its corresponding resolution, contrast, and the data of each monitoring scene association information, the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment, the method further includes:

[0095] Step S206: Receive and process the number of frames in the current data segment containing the target person and the multiple sequence numbers corresponding to the multiple frames containing the target person in the AI ​​data to be verified;

[0096] The process of receiving and processing the number of frames in the current data segment containing the target person and the multiple sequence numbers corresponding to the multiple frames containing the target person includes: synchronously displaying the number of frames in the current data segment containing the target person and the multiple sequence numbers corresponding to the multiple frames containing the target person.

[0097] For example, when the number of frames showing the target person in the current data segment is non-zero, such as 50, the number of frames showing the target person is 5, such as frames 50-75 and frames 2000-2025. The multiple sequences corresponding to the multiple frames showing the target person are 50-75 and 2000-2025 respectively.

[0098] The process of receiving and processing the number of frames in the current data segment containing the target person and the multiple sequence numbers corresponding to the multiple frames containing the target person also includes: wirelessly transmitting the number of frames in the current data segment containing the target person and the multiple sequence numbers corresponding to the multiple frames containing the target person into a network data packet to a remote verification service node. Example 3

[0099] Figure 4 The following is a flowchart illustrating the steps of a data collaborative verification method based on intelligent AI according to Embodiment 3 of the present invention.

[0100] like Figure 4 As shown, with Figure 2Unlike the previous implementation, after establishing a remote assistance channel with backend administrators and / or technical experts, and after using an AI data collaborative verification model to collaboratively verify the number of frames in the current data segment containing the target person and the multiple sequence numbers corresponding to the multiple frames containing the target person, based on the current data segment to be verified by AI and its corresponding resolution, contrast, and each monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment, i.e. after step S205, the method further includes:

[0101] Step S207: When the number of frames of the target person in the current data segment to be verified by AI is zero, mark the current data segment as a redundant data segment for the target person; when the number of frames of the target person in the current data segment to be verified by AI is greater than zero, mark the current data segment as a non-redundant data segment for the target person.

[0102] For example, when the number of frames showing the target person in the current data segment to be verified by AI is zero, the current data segment is marked as a redundant data segment for the target person. When the number of frames showing the target person in the current data segment to be verified by AI is greater than zero, the current data segment is marked as a non-redundant data segment for the target person. This includes: when the current data segment is marked as a redundant data segment for the target person, the current data segment may also be marked as a non-redundant data segment for other people who are not the target person. Example 4

[0103] Figure 5 The following is a flowchart illustrating the steps of a data collaborative verification method based on intelligent AI according to Embodiment 4 of the present invention.

[0104] like Figure 5 As shown, with Figure 2 Unlike the previous embodiment, after performing each learning action on the multilayer feedforward neural network to obtain the multilayer feedforward neural network after performing each learning action and using it as an AI data collaborative verification model, that is, after step S204, the method further includes:

[0105] Step S208: Receive the AI ​​data collaborative verification model and complete the model storage of the AI ​​data collaborative verification model by storing the various model parameters of the AI ​​data collaborative verification model;

[0106] The process of receiving AI data collaborative verification models and storing the AI ​​data collaborative verification models by storing various model parameters includes using TF storage chips, MMC storage chips, or FLASH storage chips to receive and store AI data collaborative verification models.

[0107] Specifically, multiple different physical storage addresses can be selected to store the various model parameters of the AI ​​data collaborative verification model respectively.

[0108] Next, the various method embodiments of the present invention will be described in detail.

[0109] In the data collaborative verification method based on intelligent AI according to various method embodiments of the present invention:

[0110] The AI ​​data collaborative verification model is based on the current data segment to be verified by AI, its corresponding resolution, contrast, and the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment. It collaboratively verifies the number of frames in the current data segment in which the target person appears, as well as the multiple serial numbers corresponding to the multiple frames in which the target person appears. The model inputs the current data segment to be verified by AI, its corresponding resolution, contrast, and the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment in the current data segment to be verified by AI in parallel into the AI ​​data collaborative verification model.

[0111] Specifically, the resolution corresponding to the current data segment to be verified by AI can be 1280 pixels × 720 pixels, 1920 pixels × 1080 pixels, 3840 pixels × 2160 pixels, etc.

[0112] The AI ​​data collaborative verification model is used to verify the current data segment to be verified by AI, its corresponding resolution, contrast, and the data of each monitoring scene association information, the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment. It also verifies the number of frames in the current data segment in which the target person appears and the multiple serial numbers corresponding to the multiple frames in which the target person appears. The AI ​​data collaborative verification model is run to obtain the number of frames in the current data segment in which the target person appears and the multiple serial numbers corresponding to the multiple frames in which the target person appears.

[0113] Among them, the current data segment to be verified by AI data and its corresponding resolution, contrast, and data of each monitoring scene association information, the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all screen frames in the current data segment to be verified by AI data are represented in binary numerical form when they are input into the AI ​​data collaborative verification model in parallel.

[0114] Specifically, programmable logic devices can be selected to complete the parallel input of the current data segment to be verified by AI, its corresponding resolution, contrast, and data of each monitoring scene association information, the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all screen frames in the current data segment to be verified by AI.

[0115] Among them, the number of frames in the current data segment where the target person appears and the multiple serial numbers corresponding to the multiple frames of the target person appearing, output by the AI ​​data collaborative verification model, are represented in binary numerical form.

[0116] In addition, the current data segment to be verified by AI data and its corresponding resolution, contrast, and data of each monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment to be verified by AI data are represented in binary numerical form when they are input into the AI ​​data collaborative verification model in parallel. The data of the current data segment to be verified by AI data and the data of the digital file corresponding to the baseline body shape pattern of the target person are both binary bitstreams.

[0117] And in the data collaborative verification method based on intelligent AI according to various method embodiments of the present invention:

[0118] The numerical trend of the number of times the learned action is positively correlated with the total number of pixels in the baseline body shape pattern includes: a numerical mapping formula is used to represent the numerical mapping relationship of the number of times the learned action is positively correlated with the total number of pixels in the baseline body shape pattern.

[0119] Specifically, the MATLAB toolbox can be used to test and simulate the data processing of the numerical mapping relationship, which uses a numerical mapping formula to represent the positive correlation between the number of learned actions and the total number of pixels in the baseline body shape pattern.

[0120] The numerical mapping relationship, which uses a numerical mapping formula to represent the positive correlation between the number of learning actions and the total number of pixels in the baseline body shape pattern, includes: in the numerical mapping formula, the total number of pixels in the baseline body shape pattern is the input value of the numerical mapping formula.

[0121] Furthermore, the numerical mapping relationship, which uses a numerical mapping formula to represent the positive correlation between the number of learning actions and the total number of pixels in the baseline body shape pattern, also includes: in the numerical mapping formula, the number of learning actions performed by the multilayer feedforward neural network corresponding to the total number of pixels in the baseline body shape pattern is the output value of the numerical mapping formula. Example 5

[0122] Figure 6 This is a schematic diagram of the structure of a data collaborative verification system based on intelligent AI, as shown in Embodiment 5 of the present invention.

[0123] like Figure 6 As shown, the data collaborative verification system based on intelligent AI includes a memory and multiple processors. The memory stores a computer program, which is configured to be executed by the multiple processors to complete the following steps:

[0124] Step 201: Obtain the current data segment to be verified by AI. The current data segment is obtained by concatenating the data of multiple digital files corresponding to multiple frames with the same resolution and contrast. The multiple digital files all adopt the first image file format and the multiple frames are assigned their respective serial numbers according to the order of concatenation.

[0125] Specifically, image file format refers to the storage format of the underlying data of a computer image. Common image file formats include bmp, jpeg, png, tif, gif, pcx, tga, exif, fpx, svg, psd, cdr, pcd, dxf, ufo, eps, ai, raw, WMF, webp, avif, apng, etc.

[0126] More specifically, BMP format, or bitmap format, is a hardware-independent image file format that is widely used. It uses a bitmap storage format and does not use any other compression except for the selectable image depth. The image depth of BMP file format can be 1 bit, 4 bit, 8 bit, or 24 bit. When storing data in BMP file format, the image is scanned in a left-to-right, bottom-to-top order.

[0127] More specifically, the JPEG format, also known as the Joint Photo Experts Group format, is the most common image format. It is a lossy compression format that can compress images into a smaller storage space. It uses lossy compression to remove redundant image data, achieving a very high compression ratio while displaying rich and vivid images. In other words, it can obtain good image quality with minimal disk space. The JPEG file format mainly compresses high-frequency information and retains color information well. It is suitable for use on the Internet, can reduce image transmission time, can support 24-bit true color, and is also widely used for images that require continuous tones.

[0128] Step 202: Obtain the data of the digital file corresponding to the baseline body shape pattern of the target person. The digital file corresponding to the baseline body shape pattern of the target person adopts the second image file format.

[0129] Clearly, this enables personnel detection and localization in non-image recognition modes across file formats because the baseline body shape pattern of the target personnel participating in data collaborative verification and the image in the current data segment to be verified by AI can use different image file formats.

[0130] For example, the baseline body shape image of the target person can be in jpeg format, while the image in the current data segment to be verified by AI can be in bmp format;

[0131] Step 203: Obtain the monitoring scene association information corresponding to the current data segment to be verified by AI;

[0132] Specifically, the monitoring scene association information corresponding to the current data segment to be verified by AI includes the camera parameters of the camera device that performs the monitoring screen acquisition, as well as the actual scene information of the monitoring scene;

[0133] Step 204: Perform each learning action on the multilayer feedforward neural network to obtain the multilayer feedforward neural network after each learning action is performed, and use it as an AI data collaborative verification model;

[0134] Specifically, numerical simulation mode can be used to complete the testing and simulation of the model building process of performing each learning action on the multilayer feedforward neural network to obtain the multilayer feedforward neural network after performing each learning action, and to serve as the model building process of the AI ​​data collaborative verification model.

[0135] Step 205: Using the AI ​​data collaborative verification model, based on the current data segment to be verified by AI and its corresponding resolution, contrast, and the data of each monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment, collaboratively verify the number of frames in the current data segment in which the target person appears, and the multiple serial numbers corresponding to the multiple frames in which the target person appears.

[0136] Among them, the number of times the learning action is performed follows a positive correlation with the total number of pixels in the baseline body shape pattern.

[0137] For example, the total number of pixels in the baseline body shape pattern is 1280 pixels × 720 pixels, and the number of times the learning action is performed is 800; the total number of pixels in the baseline body shape pattern is 1920 pixels × 1080 pixels, and the number of times the learning action is performed is 900; the total number of pixels in the baseline body shape pattern is 3840 pixels × 2160 pixels, and the number of times the learning action is performed is 1000, and so on.

[0138] Among them, the number of frames in a multi-frame scene is greater than or equal to a preset frame number threshold;

[0139] Specifically, the preset frame rate threshold is relatively large, for example, 100,000 frames. In this way, the multi-frame images used for search personnel can be determined to be a massive number of frames.

[0140] In each learning action performed on the multilayer feedforward neural network, the number of frames in a known data segment showing a certain person and the multiple sequence numbers corresponding to the multiple frames showing the person are used as the output content of the multilayer feedforward neural network. The data segment and its corresponding resolution, contrast, and each monitoring scene association information, the data of the digital file corresponding to the reference body shape pattern of the person, the format type number corresponding to the image file format used by each frame in the data segment, the format type number corresponding to the image file format used by the reference body shape pattern of the person, and the total number of frames in the data segment are used as the input content of the multilayer feedforward neural network to complete this learning action.

[0141] Among them, obtaining the monitoring scene association information corresponding to the current data segment to be verified by AI includes: taking the scene monitoring camera from which the multiple frames in the current data segment to be verified by AI as the current scene camera, and taking the shooting frame rate, exposure time, aperture value and shutter speed of the current scene camera and the monitoring area corresponding to the monitoring scene where the current scene camera is located as the monitoring scene association information corresponding to the current data segment to be verified by AI.

[0142] Among them, the data of the digital file corresponding to the baseline body shape pattern of the target personnel is obtained. The digital file corresponding to the baseline body shape pattern of the target personnel adopts the second image file format, which includes: the resolution and contrast of each baseline body shape pattern corresponding to each person are the same.

[0143] Specifically, the definition of whether two frames have the same image standard is completed by using the same resolution and the same contrast. That is, when the resolution and contrast of two frames are the same, it is determined that the two frames have the same image standard.

[0144] The current data segment is obtained by concatenating the data of multiple digital files corresponding to multiple frames with the same resolution and contrast. The multiple digital files all adopt the first image file format, and the multiple frames are assigned their own serial numbers according to the order of concatenation. The multiple frames are assigned their own serial numbers in the current data segment according to the order of concatenation of the data of their respective multiple digital files in the current data segment.

[0145] Specifically, for multiple frames within the current data segment, different binary values, from smallest to largest, are used to represent the multiple frame sequence numbers corresponding to each frame.

[0146] And wherein, the sequence number of each of the multiple frames is obtained in the current data segment according to the order in which the data of their respective multiple digital files are connected end to end in the current data segment, including: for each frame in the multiple frames, the earlier the data of its corresponding digital file participates in the connection in the current data segment, the smaller its sequence number in the current data segment;

[0147] like Figure 6 As shown, for example, S processors are given, where S is a natural number greater than or equal to 1. Example 6

[0148] Figure 7 This is a schematic diagram of the structure of a data collaborative verification system based on intelligent AI, as shown in Embodiment 6 of the present invention.

[0149] like Figure 7 As shown, the data collaborative verification system based on intelligent AI includes the following components:

[0150] The first capturing device is used to acquire the current data segment to be verified by AI. The current data segment is obtained by connecting the data of multiple digital files corresponding to multiple frames with the same resolution and the same contrast. The multiple digital files all adopt the first image file format and the multiple frames are assigned their respective serial numbers according to the order of connection.

[0151] Specifically, image file format refers to the storage format of the underlying data of a computer image. Common image file formats include bmp, jpeg, png, tif, gif, pcx, tga, exif, fpx, svg, psd, cdr, pcd, dxf, ufo, eps, ai, raw, WMF, webp, avif, apng, etc.

[0152] More specifically, BMP format, or bitmap format, is a hardware-independent image file format that is widely used. It uses a bitmap storage format and does not use any other compression except for the selectable image depth. The image depth of BMP file format can be 1 bit, 4 bit, 8 bit, or 24 bit. When storing data in BMP file format, the image is scanned in a left-to-right, bottom-to-top order.

[0153] More specifically, the JPEG format, also known as the Joint Photo Experts Group format, is the most common image format. It is a lossy compression format that can compress images into a smaller storage space. It uses lossy compression to remove redundant image data, achieving a very high compression ratio while displaying rich and vivid images. In other words, it can obtain good image quality with minimal disk space. The JPEG file format mainly compresses high-frequency information and retains color information well. It is suitable for use on the Internet, can reduce image transmission time, can support 24-bit true color, and is also widely used for images that require continuous tones.

[0154] The second capturing device is used to acquire data of a digital file corresponding to the baseline body shape pattern of the target person. The digital file corresponding to the baseline body shape pattern of the target person adopts a second image file format.

[0155] Clearly, this enables personnel detection and localization in non-image recognition modes across file formats because the baseline body shape pattern of the target personnel participating in data collaborative verification and the image in the current data segment to be verified by AI can use different image file formats.

[0156] For example, the baseline body shape image of the target person can be in jpeg format, while the image in the current data segment to be verified by AI can be in bmp format;

[0157] The third capture device is used to acquire the monitoring scene association information corresponding to the current data segment to be verified by AI.

[0158] Specifically, the monitoring scene association information corresponding to the current data segment to be verified by AI includes the camera parameters of the camera device that performs the monitoring screen acquisition, as well as the actual scene information of the monitoring scene;

[0159] Machine learning devices are used to perform various learning actions on a multi-layer feedforward neural network to obtain a multi-layer feedforward neural network after each learning action is performed, and to serve as an AI data collaborative verification model.

[0160] Specifically, numerical simulation mode can be used to complete the testing and simulation of the model building process of performing each learning action on the multilayer feedforward neural network to obtain the multilayer feedforward neural network after performing each learning action, and to serve as the model building process of the AI ​​data collaborative verification model.

[0161] The collaborative verification device is connected to the first capture device, the second capture device, and the third capture device respectively. It is used to use the AI ​​data collaborative verification model to verify the current data segment to be verified by AI, its corresponding resolution, contrast, and the data of each monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment. It also verifies the number of frames in the current data segment in which the target person appears and the multiple serial numbers corresponding to the multiple frames in which the target person appears.

[0162] Among them, the number of times the learning action is performed follows a positive correlation with the total number of pixels in the baseline body shape pattern.

[0163] For example, the total number of pixels in the baseline body shape pattern is 1280 pixels × 720 pixels, and the number of times the learning action is performed is 800; the total number of pixels in the baseline body shape pattern is 1920 pixels × 1080 pixels, and the number of times the learning action is performed is 900; the total number of pixels in the baseline body shape pattern is 3840 pixels × 2160 pixels, and the number of times the learning action is performed is 1000, and so on.

[0164] Among them, the number of frames in a multi-frame scene is greater than or equal to a preset frame number threshold;

[0165] Specifically, the preset frame rate threshold is relatively large, for example, 100,000 frames. In this way, the multi-frame images used for search personnel can be determined to be a massive number of frames.

[0166] In each learning action performed on the multilayer feedforward neural network, the number of frames in a known data segment showing a certain person and the multiple sequence numbers corresponding to the multiple frames showing the person are used as the output content of the multilayer feedforward neural network. The data segment and its corresponding resolution, contrast, and each monitoring scene association information, the data of the digital file corresponding to the reference body shape pattern of the person, the format type number corresponding to the image file format used by each frame in the data segment, the format type number corresponding to the image file format used by the reference body shape pattern of the person, and the total number of frames in the data segment are used as the input content of the multilayer feedforward neural network to complete this learning action.

[0167] Among them, obtaining the monitoring scene association information corresponding to the current data segment to be verified by AI includes: taking the scene monitoring camera from which the multiple frames in the current data segment to be verified by AI as the current scene camera, and taking the shooting frame rate, exposure time, aperture value and shutter speed of the current scene camera and the monitoring area corresponding to the monitoring scene where the current scene camera is located as the monitoring scene association information corresponding to the current data segment to be verified by AI.

[0168] Among them, the data of the digital file corresponding to the baseline body shape pattern of the target personnel is obtained. The digital file corresponding to the baseline body shape pattern of the target personnel adopts the second image file format, which includes: the resolution and contrast of each baseline body shape pattern corresponding to each person are the same.

[0169] Specifically, the definition of whether two frames have the same image standard is completed by using the same resolution and the same contrast. That is, when the resolution and contrast of two frames are the same, it is determined that the two frames have the same image standard.

[0170] The current data segment is obtained by concatenating the data of multiple digital files corresponding to multiple frames with the same resolution and contrast. The multiple digital files all adopt the first image file format, and the multiple frames are assigned their own serial numbers according to the order of concatenation. The multiple frames are assigned their own serial numbers in the current data segment according to the order of concatenation of the data of their respective multiple digital files in the current data segment.

[0171] Specifically, for multiple frames within the current data segment, different binary values, from smallest to largest, are used to represent the multiple frame sequence numbers corresponding to each frame.

[0172] And wherein, the sequence number of each of the multiple frames is obtained in the current data segment according to the order in which the data of their respective multiple digital files are connected end to end in the current data segment, including: for each frame in the multiple frames, the earlier the data of its corresponding digital file participates in the connection in the current data segment, the smaller its sequence number in the current data segment.

[0173] Furthermore, the present invention may also reference the following technical contents to further demonstrate the outstanding substantial progress of the present invention:

[0174] Performing learning actions on a multi-layer feedforward neural network to obtain a multi-layer feedforward neural network after each learning action and using it as an AI data collaborative verification model includes: the multi-layer feedforward neural network includes a single input layer, multiple hidden layers and a single output layer. In the multi-layer feedforward neural network, the signal propagates unidirectionally and each layer of neurons is only fully connected to the adjacent layer. Feature transformation is achieved through layer-by-layer cascading.

[0175] The multilayer feedforward neural network includes a single input layer, multiple hidden layers, and a single output layer. In the multilayer feedforward neural network, the signal propagates unidirectionally and each layer's neurons are only fully connected to the adjacent layers. Feature transformation is achieved through layer-by-layer cascading, including: the number of hidden layers in the multilayer feedforward neural network follows the numerical trend of the contrast of the baseline body shape pattern.

[0176] The variation trend of the number of hidden layers in the multilayer feedforward neural network following the numerical change trend of the contrast of the reference body shape pattern includes: the numerical change trend of the contrast of the reference body shape pattern presents a curve of change of a first linear function, and the numerical change trend of the number of hidden layers in the multilayer feedforward neural network presents a curve of change of a second linear function.

[0177] For example, the numerical trend of the contrast of the reference body shape pattern is presented as a curve of the first linear function, and the numerical trend of the number of hidden layers in the multilayer feedforward neural network is presented as a curve of the second linear function. This includes: the first programmable logic device and the second programmable logic device can be selected to implement the testing and simulation of the first linear function and the second linear function respectively.

[0178] Furthermore, the number of hidden layers in the multilayer feedforward neural network following the numerical trend of the contrast of the reference body shape pattern also includes: the curve of the first linear function and the curve of the second linear function have the same curvature.

[0179] The above description is merely a specific implementation of the embodiments of this application, but the protection scope of the embodiments of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the embodiments of this application should be covered within the protection scope of the embodiments of this application. Therefore, the protection scope of the embodiments of this application should be determined by the protection scope of the claims.

Claims

1. A data collaborative verification method based on intelligent AI, characterized in that, The method includes: The current data segment to be verified by AI is obtained. The current data segment is obtained by concatenating the data of multiple digital files corresponding to multiple frames with the same resolution and the same contrast. The multiple digital files all adopt the first image file format and the multiple frames are assigned their respective serial numbers according to the order of concatenation. The data of the digital file corresponding to the baseline body shape pattern of the target person is obtained, and the digital file corresponding to the baseline body shape pattern of the target person adopts the second image file format; Obtain the monitoring scene association information corresponding to the current data segment to be verified by AI; Perform each learning action on the multilayer feedforward neural network to obtain the multilayer feedforward neural network after each learning action is completed, and use it as an AI data collaborative verification model. The AI ​​data collaborative verification model is adopted to verify the current data segment to be verified by AI, its corresponding resolution, contrast, and the data of each monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment. The model is used to collaboratively verify the number of frames in the current data segment in which the target person appears, as well as the multiple serial numbers corresponding to the multiple frames in which the target person appears. Among them, the number of times the learning action is performed follows a positive correlation with the total number of pixels in the baseline body shape pattern. Among them, the number of frames in a multi-frame image is greater than or equal to a preset frame number threshold.

2. The data collaborative verification method based on intelligent AI as described in claim 1, characterized in that: In each learning action performed on the multilayer feedforward neural network, the number of frames in a known data segment showing a certain person and the multiple sequence numbers corresponding to the multiple frames showing the person are used as the output content of the multilayer feedforward neural network. The data of the data segment and its corresponding resolution, contrast, and each monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the person, the format type number corresponding to the image file format used by each frame in the data segment, the format type number corresponding to the image file format used by the baseline body shape pattern of the person, and the total number of frames in the data segment are used as the input content of the multilayer feedforward neural network to complete this learning action. Among them, obtaining the monitoring scene association information corresponding to the current data segment to be verified by AI includes: taking the scene monitoring camera from which the multiple frames in the current data segment to be verified by AI as the current scene camera, and taking the shooting frame rate, exposure time, aperture value and shutter speed of the current scene camera and the monitoring area corresponding to the monitoring scene where the current scene camera is located as the monitoring scene association information corresponding to the current data segment to be verified by AI. Among them, the data of the digital file corresponding to the baseline body shape pattern of the target personnel is obtained. The digital file corresponding to the baseline body shape pattern of the target personnel adopts the second image file format, which includes: the resolution and contrast of each baseline body shape pattern corresponding to each person are the same.

3. The data collaborative verification method based on intelligent AI as described in claim 2, characterized in that: The current data segment is obtained by concatenating the data of multiple digital files corresponding to multiple frames with the same resolution and contrast. The multiple digital files all adopt the first image file format, and the multiple frames are assigned their own serial numbers according to the order of concatenation. The multiple frames are assigned their own serial numbers in the current data segment according to the order of concatenation of the data of their respective multiple digital files in the current data segment. The process of obtaining the sequence number of each of the multiple frames in the current data segment based on the order in which the data of their respective multiple digital files are connected end to end in the current data segment includes: for each frame in the multiple frames, the earlier the data of its corresponding digital file participates in the connection in the current data segment, the smaller its sequence number in the current data segment.

4. The data collaborative verification method based on intelligent AI as described in claim 3, characterized in that, After employing an AI data collaborative verification model to collaboratively verify the current data segment to be verified by AI, based on its corresponding resolution, contrast, and the associated information of each monitoring scene, the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment, the method further includes: Receive and process the number of frames in the current data segment where the target person appears, as well as the multiple sequence numbers corresponding to the multiple frames in which the target person appears; The process of receiving and processing the number of frames in the current data segment containing the target person and the multiple sequence numbers corresponding to the multiple frames containing the target person includes: synchronously displaying the number of frames in the current data segment containing the target person and the multiple sequence numbers corresponding to the multiple frames containing the target person. The process of receiving and processing the number of frames in the current data segment containing the target person and the multiple sequence numbers corresponding to the multiple frames containing the target person also includes: wirelessly transmitting the number of frames in the current data segment containing the target person and the multiple sequence numbers corresponding to the multiple frames containing the target person into a network data packet and then wirelessly transmitting them to the remote verification service node.

5. The data collaborative verification method based on intelligent AI as described in claim 3, characterized in that, After employing an AI data collaborative verification model to collaboratively verify the current data segment to be verified by AI, based on its corresponding resolution, contrast, and the associated information of each monitoring scene, the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment, the method further includes: When the number of frames showing the target person in the current data segment to be verified by AI is zero, the current data segment is marked as a redundant data segment for the target person. When the number of frames showing the target person in the current data segment to be verified by AI is greater than zero, the current data segment is marked as a non-redundant data segment for the target person.

6. The data collaborative verification method based on intelligent AI as described in claim 3, characterized in that, After performing various learning actions on the multilayer feedforward neural network to obtain the multilayer feedforward neural network after performing various learning actions, and using it as an AI data collaborative verification model, the method further includes: Receive the AI ​​data collaborative verification model and complete the model storage of the AI ​​data collaborative verification model by storing the various model parameters of the AI ​​data collaborative verification model; The process of receiving and storing AI data collaborative verification models, and storing the AI ​​data collaborative verification models by storing the various model parameters of the AI ​​data collaborative verification models, includes using TF storage chips, MMC storage chips, or FLASH storage chips to receive and store AI data collaborative verification models.

7. The data collaborative verification method based on intelligent AI as described in any one of claims 3-6, characterized in that: The AI ​​data collaborative verification model is based on the current data segment to be verified by AI, its corresponding resolution, contrast, and the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment. It collaboratively verifies the number of frames in the current data segment in which the target person appears, as well as the multiple serial numbers corresponding to the multiple frames in which the target person appears. The model inputs the current data segment to be verified by AI, its corresponding resolution, contrast, and the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment in the current data segment to be verified by AI in parallel into the AI ​​data collaborative verification model. The AI ​​data collaborative verification model is used to verify the current data segment to be verified by AI, its corresponding resolution, contrast, and the data of each monitoring scene association information, the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment. It also verifies the number of frames in the current data segment in which the target person appears and the multiple serial numbers corresponding to the multiple frames in which the target person appears. The AI ​​data collaborative verification model is run to obtain the number of frames in the current data segment in which the target person appears and the multiple serial numbers corresponding to the multiple frames in which the target person appears. Among them, the current data segment to be verified by AI data and its corresponding resolution, contrast, and data of each monitoring scene association information, the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all screen frames in the current data segment to be verified by AI data are represented in binary numerical form when they are input into the AI ​​data collaborative verification model in parallel. Among them, the number of frames in the current data segment where the target person appears and the multiple serial numbers corresponding to the multiple frames of the target person appearing, output by the AI ​​data collaborative verification model, are represented in binary numerical form. Among them, the current data segment to be verified by AI data and its corresponding resolution, contrast, and data of each monitoring scene association information, the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment to be verified by AI data are represented in binary numerical form when they are input into the AI ​​data collaborative verification model in parallel. The data of the current data segment to be verified by AI data and the digital file corresponding to the baseline body shape pattern of the target person are both binary bitstreams.

8. The data collaborative verification method based on intelligent AI as described in any one of claims 3-6, characterized in that: The numerical trend of the number of times the learned action is positively correlated with the total number of pixels in the baseline body shape pattern includes: a numerical mapping formula is used to represent the numerical mapping relationship of the number of times the learned action is positively correlated with the total number of pixels in the baseline body shape pattern. The numerical mapping relationship, which uses a numerical mapping formula to represent the positive correlation between the number of learning actions and the total number of pixels in the baseline body shape pattern, includes: in the numerical mapping formula, the total number of pixels in the baseline body shape pattern is the input value of the numerical mapping formula. The numerical mapping relationship, which uses a numerical mapping formula to represent the positive correlation between the number of learning actions and the total number of pixels in the baseline body shape pattern, further includes: in the numerical mapping formula, the number of learning actions performed by the multilayer feedforward neural network corresponding to the total number of pixels in the baseline body shape pattern is the output value of the numerical mapping formula.

9. A data collaborative verification system based on intelligent AI, characterized in that, The system includes a memory and multiple processors, the memory storing a computer program, characterized in that the computer program is configured to be executed by the multiple processors to complete the following steps: The current data segment to be verified by AI is obtained. The current data segment is obtained by concatenating the data of multiple digital files corresponding to multiple frames with the same resolution and the same contrast. The multiple digital files all adopt the first image file format and the multiple frames are assigned their respective serial numbers according to the order of concatenation. The data of the digital file corresponding to the baseline body shape pattern of the target person is obtained, and the digital file corresponding to the baseline body shape pattern of the target person adopts the second image file format; Obtain the monitoring scene association information corresponding to the current data segment to be verified by AI; Perform each learning action on the multilayer feedforward neural network to obtain the multilayer feedforward neural network after each learning action is completed, and use it as an AI data collaborative verification model. The AI ​​data collaborative verification model is adopted to verify the current data segment to be verified by AI, its corresponding resolution, contrast, and the data of each monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment. The model is used to collaboratively verify the number of frames in the current data segment in which the target person appears, as well as the multiple serial numbers corresponding to the multiple frames in which the target person appears. Among them, the number of times the learning action is performed follows a positive correlation with the total number of pixels in the baseline body shape pattern. Among them, the number of frames in a multi-frame image is greater than or equal to a preset frame number threshold.

10. A data collaborative verification system based on intelligent AI, characterized in that, The system includes: The first capturing device is used to acquire the current data segment to be verified by AI. The current data segment is obtained by connecting the data of multiple digital files corresponding to multiple frames with the same resolution and the same contrast. The multiple digital files all adopt the first image file format and the multiple frames are assigned their respective serial numbers according to the order of connection. The second capturing device is used to acquire data of a digital file corresponding to the baseline body shape pattern of the target person. The digital file corresponding to the baseline body shape pattern of the target person adopts a second image file format. The third capture device is used to acquire the monitoring scene association information corresponding to the current data segment to be verified by AI. Machine learning devices are used to perform various learning actions on a multi-layer feedforward neural network to obtain a multi-layer feedforward neural network after each learning action is performed, and to serve as an AI data collaborative verification model. The collaborative verification device is connected to the first capture device, the second capture device, and the third capture device respectively. It is used to use the AI ​​data collaborative verification model to verify the current data segment to be verified by AI, its corresponding resolution, contrast, and the data of each monitoring scene association information, the data of the digital file corresponding to the baseline body shape pattern of the target person, the format type number of the first image file format and the second image file format, and the number of all frames in the current data segment. It also verifies the number of frames in the current data segment in which the target person appears and the multiple serial numbers corresponding to the multiple frames in which the target person appears. Among them, the number of times the learning action is performed follows a positive correlation with the total number of pixels in the baseline body shape pattern. Among them, the number of frames in a multi-frame image is greater than or equal to a preset frame number threshold.

Images