Video stream storage system, video stream retrieval matching system and corresponding methods

Abstract

The invention relates to the technical field of computer vision and image processing, and provides a video stream storage system, a video stream retrieval matching system and a corresponding method.A video collection device with high resolution, low illumination and wide dynamic range is selected to collect video stream data of a field, and a large-capacity hard disk and cloud storage backup are configured; an advanced storage algorithm is adopted to carry out distributed storage and compression on the data, so that the storage efficiency and the space utilization rate are improved; and an effective video indexing mechanism is established, so that the video clips of a specific time period, a specific area and a specific person can be quickly positioned and retrieved. And quick retrieval is performed through timestamp information.

Description

Technical Field

[0001] The present invention relates to the technical field of computer vision and image processing, and more specifically, to a video stream storage system, a video stream retrieval and matching system, and corresponding methods. Background Art

[0002] The amount of data in high-definition video streams is huge, and long-term storage of these videos requires a lot of storage resources. For example, an ordinary high-definition camera generates tens to hundreds of GB of data every day. For large enterprises, the data storage costs generated by thousands of cameras are a huge burden.

[0003] When it is necessary to review videos, it is very difficult to find video clips related to a specific time period, a specific area, or a specific person from a large amount of video data without an efficient retrieval mechanism. For example, in a surveillance system of a large shopping mall, to find the activity track of a suspicious person on a certain day, it may take a lot of manual time to screen through numerous video files. Summary of the invention

[0004] The present invention aims at the technical problems existing in the prior art and provides a video stream storage system, a video stream retrieval and matching system and a corresponding method to solve the problems of high video storage cost and low retrieval efficiency in the prior art.

[0005] According to a first aspect of the present invention, a video stream storage system is provided, comprising:

[0006] Front-end equipment layer, data transmission layer, back-end processing layer and user terminal layer;

[0007] The front-end device layer includes a video acquisition device for collecting case scene video stream data;

[0008] The data transmission layer is used to transmit the collected case scene video stream data to the back-end processing layer based on data compression and error correction methods;

[0009] The backend processing layer includes an edge server and a cloud server. The edge server is used to decode and extract frames of the case video stream data; the cloud server is used to perform distributed storage of the case video stream data after decoding and extracting frames.

[0010] The user terminal layer includes a terminal device, and the terminal device is used to access the cloud server through client software or a web interface to view and retrieve video clips.

[0011] Based on the above technical solution, the present invention can also make the following improvements.

[0012] Optionally, the front-end device layer includes multiple video acquisition devices, each of which has different resolutions and viewing angles and can cover various key areas of the case site. The multiple video acquisition devices are used to collect video stream data of various key areas of the case site.

[0013] Optionally, the data transmission layer is used to transmit the collected case scene video stream data to the back-end processing layer based on data compression and error correction methods, including:

[0014] Compress the collected case scene video stream data;

[0015] Determine whether the size of the compressed video stream data is within a set size range. If so, transmit the compressed video stream data to the back-end processing layer; if not, re-compress the collected video stream data to within the set size range.

[0016] Optionally, the edge server is used to decode and extract frames of the case scene video stream data, including:

[0017] The edge server decodes the compressed video stream data into original video stream data, and performs frame extraction processing on the decoded original video stream data at a set time interval;

[0018] The cloud server is used to perform distributed storage on the case scene video stream data after decoding and frame extraction processing, including:

[0019] Detect whether there are living things in the video stream data screen, save the video stream data if there are living things, and delete the video stream data if there are no living things;

[0020] The saved video stream data is distributed and stored according to different areas and time periods.

[0021] According to a second aspect of the present invention, a video stream retrieval and matching system is provided, including an authentication integrated machine, a case site NVR recorder, a video acquisition device and a cloud server;

[0022] The video acquisition device records the case site monitoring video and transmits the case site monitoring video to the case site NVR recorder and the cloud server respectively;

[0023] The case site NVR recorder is used to store the case site monitoring video locally for on-site viewing;

[0024] The authentication integrated machine is used to authenticate the customer's facial information in the case site surveillance video. When the identity authentication is passed, the customer's authorized facial image is transmitted to the cloud server;

[0025] The cloud server is used to match the customer's authorized facial image in a pre-stored database to obtain customer-related information.

[0026] Optionally, the cloud server is used to match the customer's authorized facial image in a pre-stored database to obtain customer-related information, including:

[0027] According to the preset time period and area to be matched, searching the database for video clips of the corresponding time period and the corresponding area;

[0028] According to the authorized facial image of the customer, a match is made in the corresponding video clips found to obtain the matching customer information in the database.

[0029] Optionally, matching the authorized facial image of the customer in the corresponding video clips to obtain the matching customer information in the database includes:

[0030] Perform frame extraction processing on the found video clips, calculate the similarity between any frame of video image and the authorized facial image of the customer, if the similarity is greater than the similarity threshold, then any frame of video image matches the authorized facial image of the customer; otherwise, any frame of video image does not match the authorized facial image of the customer;

[0031] In the case that any frame of video image matches the authorized facial image of the customer, the recording time of any frame of video image is obtained, that is, the time when the customer visited.

[0032] Optionally, the cloud server is used to match the authorized facial image of the customer in a pre-stored database to obtain customer-related information, and further includes:

[0033] Matching tasks are set on a daily basis, and the authorized facial images of each customer of each day are matched in the database. The matching results of each customer are obtained, the matching task of the current day is completed, and the next matching task is waited for.

[0034] According to a third aspect of the present invention, there is provided a video stream storage method, comprising:

[0035] Collecting the case scene video stream data, compressing and error-correcting the case scene video stream data and transmitting it to the back-end processing layer;

[0036] Decoding and frame extraction processing are performed on the case scene video stream data at the back-end processing layer, and distributed storage is performed on the case scene video stream data after decoding and frame extraction processing;

[0037] The backend processing layer is accessed through client software or a web interface to view the retrieved video clips.

[0038] According to a fourth aspect of the present invention, a video stream retrieval and matching method is provided, comprising:

[0039] Recording case surveillance video, storing the case surveillance video locally and transmitting it to a cloud server;

[0040] The customer's facial information in the on-site surveillance video is authenticated through the authentication machine. Once the authentication is passed, the customer's authorized facial image is transmitted to the cloud server;

[0041] Based on the customer’s authorized facial image, a match is performed in the database pre-stored on the cloud server to obtain customer-related information.

[0042] The video stream storage system and video stream retrieval and matching system provided by the present invention adopt advanced storage algorithms, distributed storage, data compression, etc. in video storage to improve storage efficiency and space utilization, while reducing storage costs. In terms of data retrieval, an effective video indexing mechanism is established to facilitate rapid positioning and retrieval of video clips of a specific time period, specific area or specific person. Rapid retrieval is performed through information such as timestamps, location identifiers, and person features to improve retrieval efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A structural block diagram of a video stream storage system provided by the present invention;

[0044] Figure 2 Store flow graph for video stream;

[0045] Figure 3 A structural block diagram of a video stream retrieval and matching system provided by the present invention;

[0046] Figure 4 A flow chart of a method for storing video streams;

[0047] Figure 5 Flowchart of the matching method for retrieving video streams. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. In addition, the technical features in the various embodiments or single embodiments provided by the present invention can be arbitrarily combined with each other to form a feasible technical solution. This combination is not subject to the constraints of the sequence of steps and / or the structural composition mode, but must be based on the ability of ordinary technicians in this field to achieve. When the combination of technical solutions is contradictory or cannot be achieved, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0049] Figure 1 A structural block diagram of a video stream storage system provided by the present invention is shown in FIG. Figure 1 ,The video stream storage system includes front-end device layer, data transmission layer, back-end ,processing layer and user terminal layer.

[0050] The front-end equipment layer includes video acquisition equipment, which is used to collect video stream data from the scene;

[0051] A data transmission layer, used to transmit the collected case scene video stream data to the back-end processing layer based on data compression and error correction methods;

[0052] The backend processing layer includes an edge server and a cloud server. The edge server is used to decode and extract frames of the case video stream data; the cloud server is used to perform distributed storage of the case video stream data after decoding and extracting frames.

[0053] The user terminal layer includes a terminal device, and the terminal device is used to access the cloud server through client software or a web interface to view and retrieve video clips.

[0054] For example, the front-end device layer mainly includes video acquisition devices such as high-definition cameras, which are responsible for collecting video stream data of the case site. There are multiple cameras, each with different resolutions and viewing angles, which can cover various key areas of the case site. For example, multiple cameras are installed in each key area to capture the live video stream data in the area.

[0055] The data transmission layer uses wired Ethernet or wireless Wi-Fi network to transmit the video stream data collected by the front end to the back-end server. To ensure the stability and efficiency of data transmission, data compression and error correction technology are used.

[0056] Among them, the data transmission layer is used to transmit the collected case scene video stream data to the back-end processing layer based on data compression and error correction methods, including: the data transmission layer compresses the collected case scene video stream data; determines whether the size of the compressed video stream data is within a set size range, if so, transmits the compressed video stream data to the back-end processing layer; if not, re-compresses the collected video stream data to within the set size range.

[0057] For example, since the amount of collected on-site video stream data is relatively large, it is necessary to compress the collected video stream data during transmission. After compressing the video stream data, determine whether the size of the compressed video stream data is within a reasonable range. If the compressed video stream data is too small, it cannot be played normally later. If the collected video stream data is compressed too small, it needs to be recompressed. After compressing the video stream data to a reasonable range, it is transmitted to the back-end processing layer.

[0058] Among them, the backend processing layer mainly includes edge servers and cloud servers. The edge server decodes the compressed video stream data into original video stream data, and extracts frames from the decoded original video stream data at set time intervals;

[0059] The cloud server mainly performs data analysis on the video stream data, including detecting whether there are living things in the video stream data screen, saving the video stream data if there are living things, and deleting the video stream data if there are no living things; and distributing the saved video stream data according to different areas and time periods.

[0060] For example, since the data transmission layer compresses the collected video stream data and transmits it to the back-end processing layer, the edge server of the back-end processing layer decodes the compressed video stream data transmitted to obtain the original video stream data. Since the amount of collected video stream data is relatively large, it is necessary to extract frames from the original video stream data. Specifically, the video stream data can be extracted at fixed time intervals, such as extracting key frames therein, to obtain the video stream data after the extraction of frames.

[0061] The cloud server introduces a dynamic line analysis algorithm to the video stream data after frame extraction, divides the video stream data captured by the camera into regions, and draws the areas where the passenger flow passes. Through algorithm analysis, the video stream when there are living things passing through the area is saved, and the video is not saved when there are no living things in the area. At the same time, the recording cycle is set according to the working hours of the case site. This greatly reduces the total amount of video. It improves the efficiency of subsequent face matching and complies with laws and regulations. In order to facilitate subsequent matching, the video storage is partitioned and stored according to the date to achieve distributed storage.

[0062] When storing a large amount of video stream data, the video stream data may be classified, for example, the video streams may be classified and stored according to different time periods and different regions to facilitate subsequent search and retrieval.

[0063] See also Figure 2 , which is a flowchart of video stream data storage, mainly includes the following steps:

[0064] 1. Initial stage: The front-end device layer is responsible for collecting the video stream data of the case site.

[0065] 2. Use wired Ethernet or wireless Wi-Fi network to transmit the video stream data collected by the front end to the back-end server. To ensure the stability and efficiency of data transmission, data compression and error correction technology are used.

[0066] 3. The video analysis server (which can be understood as an edge server) performs preliminary processing on the video stream data and decodes the compressed video stream data into original image data; the frame extraction submodule extracts key frames from the video stream at a certain time interval for subsequent analysis and storage.

[0067] 4. The data collection program classifies and processes the collected videos, including local storage and cloud storage. Local storage is mainly used to temporarily store recent video stream data for quick retrieval; cloud storage is used to store video stream data for a long time, and distributed storage technology is used to ensure data security and reliability.

[0068] See also Figure 3 , which is a structural block diagram of the video stream retrieval and matching system. The video stream retrieval and matching system includes an authentication all-in-one machine, a case-site NVR recorder, a video acquisition device, and a cloud server.

[0069] The system needs to integrate multiple hardware devices and software modules such as surveillance cameras, ID card machines, edge servers and cloud servers to realize the functions of the entire system. This involves system integration technology to ensure that the various components can work together and data can be accurately transmitted and processed. At the same time, software development is also required, including writing server-side and client-side software programs to realize functions such as algorithm configuration, risk assessment management, task distribution, video storage and face matching.

[0070] Among them, the video acquisition equipment records the case site monitoring video and transmits the case site monitoring video to the case site NVR recorder and the cloud server respectively;

[0071] Case site NVR recorder, used to store case site surveillance video locally for on-site viewing;

[0072] The authentication machine is used to authenticate the customer's facial information in the on-site surveillance video. Once the identity authentication is passed, the customer's authorized facial image is transmitted to the cloud server;

[0073] The cloud server is used to match the customer's authorized facial image in a pre-stored database to obtain customer-related information.

[0074] For example, the scene includes video acquisition equipment, a case NVR recorder and an authentication integrated machine. The video acquisition equipment records the case monitoring video and sends the collected case monitoring video to the case NVR recorder and the cloud server respectively. Among them, the case NVR recorder is located at the monitoring site and mainly stores the case monitoring video locally for on-site viewing.

[0075] When a customer visits, the customer's facial information in the surveillance video is authenticated through the authentication machine, and only certain customers are allowed to visit. When the visiting customer's identity authentication is passed, the customer's authorized facial image is transmitted to the cloud server.

[0076] Among them, historical video stream data is distributedly stored in the cloud server and is stored in different time periods and different areas.

[0077] When the authentication machine transmits the customer's authorized facial image to the cloud server, the cloud server can match the customer's authorized facial image in the cloud server's database and find matching historical customer information, such as whether the customer has visited the store before and when.

[0078] Then you can set the time period and area that needs to be matched, and find the video stream data of the corresponding area and time period in the database according to the set time period and area. Then, according to the customer's authorized face information, search in the video stream data of the corresponding area and time period, find the matching face, and obtain the time when the customer visited.

[0079] Among them, according to the authorized face image of the customer, matching is performed in the corresponding video clips found to obtain the matching customer information in the database, including:

[0080] Perform frame extraction processing on the found video clips, calculate the similarity between any frame of video image and the authorized facial image of the customer, if the similarity is greater than the similarity threshold, then any frame of video image matches the authorized facial image of the customer; otherwise, any frame of video image does not match the authorized facial image of the customer;

[0081] In the case that any frame of video image matches the authorized facial image of the customer, the recording time of any frame of video image is obtained, that is, the time when the customer visited.

[0082] For example, after finding the video clips of the corresponding area and the corresponding time period in the database, the video clips are subjected to frame extraction processing to obtain multiple frames. The similarity between the authorized face image of the customer and the face in each frame is calculated. If the similarity reaches a similarity threshold, the face in the frame matches the authorized face of the customer, and the shooting time of the frame can be obtained, that is, the time when the customer has visited.

[0083] Since the video stream data stored in the database of the cloud server is segmented by time, the matching task can be performed on the surveillance videos collected on site every day.

[0084] The matching task is set in units of days, the authorized face image of each customer of each day is matched in the database, the matching result of each customer is obtained, the matching task of the current day is completed, and the next matching task is waited for.

[0085] For example, the authorized face images of customers in the surveillance videos collected on site every day are matched in the cloud database to obtain matching results. Matching the surveillance videos of one day is regarded as one matching task. If the current matching task is completed, the cloud server waits for the next matching task.

[0086] It should be noted that the matching process is a one-by-one matching of video streams, and the face analysis results in the process are only used for matching and destroyed after the task is completed. Although each matching task will be repeated, customer privacy and security are guaranteed. The optimization of the algorithm and video storage allows the entire matching result to be controlled within 24 hours. It is compatible with the needs of real estate merchants for channel risk control and customer judgment in most cases.

[0087] The cloud server is mainly responsible for algorithm configuration management and risk assessment management. It is a conventional management system. The main communication methods with the NVR recorder are http and mqtt. During the communication process, the consistency of data transmission is guaranteed by timed tasks and retry mechanisms. At the same time, the cloud is also the storage end for all result data. The cloud stores all matching results and card swiping results in order to further strengthen privacy security and compliance. The cloud has the function of deleting matching results at any time to avoid extreme cases where the algorithm matching result is not the person himself, and when the authorizer revokes the authorization for personal reasons, the relevant face images can be deleted in time.

[0088] See also Figure 4 , provides a video stream storage method, comprising:

[0089] Step 1: Collect the case scene video stream data, compress and correct the case scene video stream data and transmit it to the back-end processing layer;

[0090] Step 2, decoding and frame extraction processing are performed on the case scene video stream data at the back-end processing layer, and distributed storage is performed on the case scene video stream data after decoding and frame extraction processing;

[0091] Step 3: Access the backend processing layer through client software or a web page interface to view the retrieved video clips.

[0092] Among them, the technical features of the video stream storage method can refer to the relevant technical features of the aforementioned video stream storage system, which will not be repeated here.

[0093] See also Figure 5 , a video stream retrieval and matching method is provided, including:

[0094] Step 1', recording the case monitoring video, storing the case monitoring video locally and transmitting it to the cloud server;

[0095] Step S2', authenticate the customer's facial information in the case monitoring video through the authentication integrated machine, and when the identity authentication is passed, transmit the customer's authorized facial image to the cloud server;

[0096] Step S3', matching the authorized facial image of the customer in a database pre-stored on the cloud server to obtain customer-related information.

[0097] Among them, the technical features of the video stream retrieval and matching method can refer to the relevant technical features of the aforementioned video stream retrieval and matching system, which will not be repeated here.

[0098] The embodiment of the present invention provides a video stream storage system and a video stream retrieval and matching system, which uses high-resolution, low-illumination, and wide dynamic range video acquisition equipment to collect on-site video stream data, is equipped with a large-capacity hard disk and cloud storage backup, and uses advanced storage algorithms to perform distributed storage and compression on data to improve storage efficiency and space utilization; and establishes an effective video indexing mechanism to facilitate rapid positioning and retrieval of video clips of specific time periods, specific areas, and specific people. Rapid retrieval is performed through timestamp information.

[0099] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0100] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0101] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0102] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0103] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0104] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0105] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A video stream storage system, characterized in that: It includes front-end equipment layer, data transmission layer, back-end processing layer and user terminal layer; The front-end device layer includes a video acquisition device for collecting case scene video stream data; The data transmission layer is used to transmit the collected case scene video stream data to the back-end processing layer based on data compression and error correction methods; The backend processing layer includes an edge server and a cloud server. The edge server is used to decode and extract frames of the case video stream data; the cloud server is used to perform distributed storage of the case video stream data after decoding and extracting frames. The user terminal layer includes a terminal device, and the terminal device is used to access the cloud server through client software or a web interface to view and retrieve video clips.

2. The video stream storage system according to claim 1, characterized in that: The front-end device layer includes multiple video acquisition devices, each of which has different resolutions and viewing angles and can cover various key areas of the case. The multiple video acquisition devices are used to collect video stream data of various key areas of the case.

3. The video stream storage system according to claim 1, characterized in that: The data transmission layer is used to transmit the collected case scene video stream data to the back-end processing layer based on data compression and error correction methods, including: Compress the collected case scene video stream data; Determine whether the size of the compressed video stream data is within a set size range. If so, transmit the compressed video stream data to the back-end processing layer; if not, re-compress the collected video stream data to within the set size range.

4. The video stream storage system according to claim 1, characterized in that: The edge server is used to decode and extract frames of the case scene video stream data, including: The edge server decodes the compressed video stream data into original video stream data, and performs frame extraction processing on the decoded original video stream data at a set time interval; The cloud server is used to perform distributed storage on the case scene video stream data after decoding and frame extraction processing, including: Detect whether there are living things in the video stream data screen, save the video stream data if there are living things, and delete the video stream data if there are no living things; The saved video stream data is distributed and stored according to different areas and time periods.

5. A video stream retrieval and matching system, characterized in that: It includes an all-in-one authentication machine, a case site NVR recorder, and the video acquisition device and cloud server described in claim 1; The video acquisition device records the case site monitoring video and transmits the case site monitoring video to the case site NVR recorder and the cloud server respectively; The case site NVR recorder is used to store the case site monitoring video locally for on-site viewing; The authentication integrated machine is used to authenticate the customer's facial information in the case site surveillance video. When the identity authentication is passed, the customer's authorized facial image is transmitted to the cloud server; The cloud server is used to match the customer's authorized facial image in a pre-stored database to obtain customer-related information.

6. The video stream retrieval and matching system according to claim 5, characterized in that: The cloud server is used to match the customer's authorized facial image in a pre-stored database to obtain customer-related information, including: According to the preset time period and area to be matched, searching the database for video clips of the corresponding time period and the corresponding area; According to the authorized facial image of the customer, a match is made in the corresponding video clips found to obtain the matching customer information in the database.

7. The video stream retrieval and matching system according to claim 6, characterized in that: The step of matching the authorized facial image of the customer in the corresponding video clips to obtain the matching customer information in the database includes: Perform frame extraction processing on the found video clips, calculate the similarity between any frame of video image and the authorized facial image of the customer, if the similarity is greater than the similarity threshold, then any frame of video image matches the authorized facial image of the customer; otherwise, any frame of video image does not match the authorized facial image of the customer; In the case that any frame of video image matches the authorized facial image of the customer, the recording time of any frame of video image is obtained, that is, the time when the customer visited.

8. The video stream retrieval and matching system according to claim 5, characterized in that: The cloud server is used to match the authorized facial image of the customer in a pre-stored database to obtain customer-related information, and also includes: Matching tasks are set on a daily basis, and the authorized facial images of each customer of each day are matched in the database. The matching results of each customer are obtained, the matching task of the current day is completed, and the next matching task is waited for.

9. A video stream storage method, characterized in that: include: Collecting the case scene video stream data, compressing and error-correcting the case scene video stream data and transmitting it to the back-end processing layer; Decoding and frame extraction processing are performed on the case scene video stream data at the back-end processing layer, and distributed storage is performed on the case scene video stream data after decoding and frame extraction processing; The backend processing layer is accessed through client software or a web interface to view the retrieved video clips.

10. A video stream retrieval and matching method, characterized in that: include: Recording case surveillance video, storing the case surveillance video locally and transmitting it to a cloud server; The customer's facial information in the on-site surveillance video is authenticated through the authentication machine. Once the authentication is passed, the customer's authorized facial image is transmitted to the cloud server; Based on the customer’s authorized facial image, a match is performed in the database pre-stored on the cloud server to obtain customer-related information.

Images