Video tracing method, device, computer equipment and storage medium

By querying event location information in the blockchain and combining it with event description and time information, the problem of low efficiency in existing video tracing is solved, enabling the rapid retrieval of key content in surveillance videos.

CN114554302BActive Publication Date: 2024-11-15INDUSTRIAL AND COMMERCIAL BANK OF CHINA
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210175888.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-02-24
Publication Date
2024-11-15
Estimated Expiration
2042-02-24

AI Technical Summary

Technical Problem

Existing video tracing methods rely on manual searching, which is inefficient and cannot quickly find key content in surveillance videos.

Method used

By receiving the target event description and time information from the tracing request, the system queries the event location information in the blockchain, uses the event description and time information to query the video database, and obtains the target video clip.

Benefits of technology

It improves the efficiency of video tracing, enabling quick location and acquisition of the required video segments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114554302B_ABST
    Figure CN114554302B_ABST
Patent Text Reader

Abstract

The present application relates to a video tracing method, device, computer equipment, storage medium and computer program product. It relates to the field of blockchain technology and can be used in the field of information security. The corresponding target event location information in the blockchain is obtained by querying the target event description information and target time information carried in the tracing request, wherein the blockchain contains multiple blocks for storing different event location information corresponding to different video clips, and the event location information corresponding to the video clip includes event description information and time information. The video database is queried according to the target event location information to obtain the video clip corresponding to the target event. Compared with the traditional method of manual retrieval, this solution uses the event description information and time information to query each block in the blockchain to obtain the event location information corresponding to the event description information, and obtains the video clip of the event to be queried based on the event location information, thereby improving the efficiency of video tracing.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of blockchain technology, and in particular to a video tracing method, apparatus, computer equipment, storage medium and computer program product. Background Art

[0002] As a means of security protection, video surveillance is widely used in data centers, bank branches, vaults and other places. When an accident occurs at the monitored site, it is necessary to trace the source of the surveillance video to obtain the key content in the video. At present, the source of video surveillance is usually manually distinguished by video content and occurrence time. However, the surveillance video is long, and the manual source tracing method of distinguishing video content and occurrence time cannot quickly find the required video content.

[0003] Therefore, the current video tracing method has the defect of low retrieval efficiency. Summary of the invention

[0004] Based on this, it is necessary to provide a video tracing method, device, computer equipment, computer-readable storage medium and computer program product that can improve retrieval efficiency in response to the above technical problems.

[0005] In a first aspect, the present application provides a video source tracing method, the method comprising:

[0006] Receiving a tracing request for a target event, obtaining target event description information and target time information carried in the tracing request;

[0007] According to the target event description information and the target time information, a blockchain is queried to obtain corresponding target event location information; the blockchain includes a plurality of blocks for storing different event location information corresponding to different video clips; the event location information corresponding to the video clip includes event description information of the video clip and its time information; the video clip is obtained based on event classification of the source video, the event description information represents a description of a set behavior in the video clip, and the time information represents the time of the video clip in the video;

[0008] The video database is queried according to the target event location information to obtain a video clip corresponding to the target event.

[0009] In one embodiment, the method further comprises:

[0010] Acquire a source video, classify the source video into events, obtain multiple events to be stored and multiple time information to be stored, and construct multiple blocks in the blockchain according to the multiple time information to be stored;

[0011] Acquire a target event to be stored whose event contents do not overlap from the multiple events to be stored, and generate event description information corresponding to the target event to be stored;

[0012] Event location information is obtained according to the event description information and the time information to be stored, and the event location information is stored in a block corresponding to the time information to be stored to obtain the blockchain.

[0013] In one embodiment, the event classification of the source video to obtain a plurality of events to be stored and a plurality of time information to be stored includes:

[0014] Obtain a candidate video segment containing a set behavior in the source video according to a classification behavior identifier, obtain an extended video segment of an event corresponding to the set behavior according to an integrity identifier, and obtain a video segment of an event to be detected and its corresponding timestamp based on the candidate video segment and the extended video segment;

[0015] Aligning the video clip of the event to be detected and the timestamp;

[0016] According to the aligned video clips of the event to be detected and the timestamps, the behavior probability and the completion probability corresponding to the video clips of the event to be detected are obtained; the behavior probability indicates whether the aligned video clips of the event to be detected are the events to be stored, and the completion probability indicates whether the events to be stored in the aligned video clips of the event to be detected are complete;

[0017] According to the behavior probability and the completion degree probability, the events to be stored and the time information to be stored including the complete events to be stored are determined.

[0018] In one embodiment, aligning the video clip of the event to be detected and the timestamp includes:

[0019] The timestamp is subjected to pyramid pooling processing, the processed timestamp and the video clip of the event to be detected are input into a preset recurrent neural network, and the video clip of the event to be detected and the timestamp are aligned through the preset recurrent neural network.

[0020] In one embodiment, the step of constructing a plurality of blocks in a blockchain according to the plurality of time information to be stored includes:

[0021] For each time information to be stored, determine the duration corresponding to the time information to be stored according to the timestamp contained in the time information to be stored, and generate a block of a size corresponding to the duration in the blockchain to be constructed;

[0022] The blockchain is obtained based on the multiple blocks.

[0023] In one embodiment, obtaining a target event to be stored whose event contents do not overlap from the multiple events to be stored includes:

[0024] For each frame of video image in the video clip corresponding to the event to be stored, obtaining an image frame to be detected whose similarity with other frames of video image in the video clip is less than a preset similarity threshold;

[0025] Input the image frame to be detected into a preset probability function to obtain a corresponding non-overlapping probability value, and determine the target image frame to be detected whose corresponding event content does not overlap according to the comparison result of the non-overlapping probability value and the preset hyperparameter;

[0026] According to the plurality of target to-be-detected image frames, a target to-be-stored event is obtained.

[0027] In one of the embodiments, obtaining the image frame to be detected whose similarity with other frame video images in the video segment is less than a preset similarity threshold includes:

[0028] If the frame of video image is the first frame in the video clip, according to the visual features of the frame of video image, the time information occupied by the frame of image in the video clip and the generated description information, the image frame to be detected is obtained by using a long short-term memory algorithm;

[0029] If the video image frame is not the first frame in the video clip, the image frame to be detected is obtained by a long short-term memory algorithm based on the previous image frame to be detected corresponding to the video image frame, the visual features of the video image frame, the time information occupied by the image frame in the video clip and the generated description information.

[0030] In one embodiment, generating event description information corresponding to the target event to be stored includes:

[0031] Acquire multiple sub-events included in the target event to be stored, and for each of the sub-events, determine the sub-event description corresponding to the sub-event by using a long short-term memory algorithm according to the visual features of the video clip corresponding to the sub-event and the generated sub-event description;

[0032] According to the multiple sub-event descriptions, event description information corresponding to the target event to be stored is obtained.

[0033] In one embodiment, obtaining event location information according to the event description information and the time information to be stored, storing the event location information in a block corresponding to the time information to be stored, and obtaining the blockchain includes:

[0034] The event description information and the time information to be stored are concatenated to obtain corresponding coordinates as event location information, and the event location information is stored in the block corresponding to the time information to be stored in the form of the coordinates to obtain the blockchain.

[0035] In a second aspect, the present application provides a video source tracing device, the device comprising:

[0036] A receiving module, configured to receive a tracing request for a target event, and obtain target event description information and target time information carried in the tracing request;

[0037] A tracing module is used to query the blockchain according to the target event description information and the target time information to obtain the corresponding target event location information; the blockchain contains a plurality of blocks for storing different event location information corresponding to different video clips; the event location information corresponding to the video clip includes the event description information of the video clip and its time information; the video clip is obtained based on event classification of the source video, the event description information represents the description of the set behavior in the video clip, and the time information represents the time of the video clip in the video;

[0038] The query module is used to query the video database according to the target event location information to obtain the video clip corresponding to the target event.

[0039] In a third aspect, the present application further provides a computer device, wherein the computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0040] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the above method when executed by a processor.

[0041] In a fifth aspect, the present application further provides a computer program product, wherein the computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0042] The above-mentioned video tracing method, device, computer equipment, storage medium and computer program product obtain the corresponding target event location information in the blockchain by querying the blockchain according to the target event description information and target time information carried in the tracing request when receiving the tracing request, wherein the blockchain contains multiple blocks for storing different event location information corresponding to different video clips obtained by classifying the source video, and the event location information corresponding to the video clip includes the event description information obtained by describing the set behavior in the video clip and the time information of the above video clip in the video. Thus, the video database is queried according to the target event location information to obtain the video clip corresponding to the target event. Compared with the traditional method of manual retrieval, this solution uses the event description information and time information to query each block in the blockchain to obtain the event location information corresponding to the event description information, so that the video clip of the event to be queried can be obtained based on the event location information, thereby improving the efficiency of video tracing. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A diagram showing an application environment of a video source tracing method in an embodiment;

[0044] Figure 2 A schematic diagram of a process of tracing the source of a video in one embodiment;

[0045] Figure 3 A schematic diagram of a process flow of an event classification step in an embodiment;

[0046] Figure 4 A schematic diagram of a flow chart of a video source tracing method in another embodiment;

[0047] Figure 5 A schematic diagram of a flow chart of a video source tracing method in another embodiment;

[0048] Figure 6 is a structural block diagram of a video source tracing device in one embodiment;

[0049] Figure 7 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0051] The video source tracing method provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 in the blockchain through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers. The terminal 102 can receive a traceability request for the target event triggered by the user, and query the blockchain based on the target event description information and target time information in the traceability request to obtain the corresponding target event location information, so that the video clip corresponding to the target event can be obtained by querying the video database based on the event description information and time information in the target event location information obtained by the query. Among them, the terminal 102 can be, but is not limited to, various personal computers, laptops, smart phones, tablet computers and portable wearable devices. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server 104 in the blockchain can be implemented with an independent server or a server cluster consisting of multiple servers.

[0052] In one embodiment, Figure 2 As shown, a video tracing method is provided, which is applied to Figure 1 The terminal in is used as an example to illustrate, including the following steps:

[0053] Step S202: receiving a tracing request for a target event, and obtaining target event description information and target time information carried in the tracing request.

[0054] Among them, the target event can be an event that the user needs to trace the source, and the event can be recorded by the image acquisition device to form a corresponding video clip. The terminal 102 can receive the user's traceability request for the target event, and obtain the target event description information and target time information carried in the traceability request. Among them, the target event description information can be the user's description of the target event that needs to be traced, and the description can be in the form of a natural language; the target time information can be the occurrence time of the target event in the source video. The terminal 102 can trace the target event based on the user's natural language description and time information. For example, the video clip corresponding to the target event is queried. In addition, in some embodiments, the information input by the user can also include the location information of the above-mentioned target event. For example, if the target event is that the user withdraws money from a bank, the location where the target event occurs can be the information of a certain branch, and the terminal 102 can store the traceability information of the target event according to the location where the target event occurs.

[0055] Step S204, query the blockchain according to the target event description information and the target time information to obtain the corresponding target event location information; the blockchain contains multiple blocks for storing different event location information corresponding to different video clips; the event location information corresponding to the video clip includes the event description information of the video clip and its time information; the video clip is obtained based on event classification of the source video, the event description information represents the description of the set behavior in the video clip, and the time information represents the time of the video clip in the video.

[0056] Among them, the blockchain can be a proof blockchain, and the blockchain contains multiple blocks for storing different event location information corresponding to different video clips. That is, each block in the blockchain can store event location information corresponding to a video clip. The size of each block can be determined according to the duration of the video clip. The terminal 102 can query the blockchain according to the target event description information and target time information in the traceability request triggered by the above user, and obtain the target event location information corresponding to the target event description information and the target time information, so that the terminal 102 can obtain the event description information and time information in the target event location information. That is, the user can initiate a traceability request in a natural language description manner, and the terminal 102 can query the blockchain based on the user's natural language description information, obtain the corresponding target block, and obtain the target event location information from the target block. Among them, the event location information corresponding to the above video clip includes the event description information of the video clip and its time information, the above video clip can be a video clip corresponding to a certain event, the video clip can be a part of the source video, and the source video can be the most complete original video, such as a complete video recorded by a certain image acquisition device. The terminal 102 can obtain multiple video clips by performing a push-type event classification on the source video, that is, each video clip can represent an event. The event includes event description information and time information. The event description information may be a description of the set behavior in the corresponding video clip. For example, if the video clip is a user's behavior of withdrawing money, the event description information may be a description of a series of user behaviors from the time the user appears in the video screen to the time the user leaves the video screen. The description may be in the form of natural language, so that when tracing the source, the corresponding event location information can be queried only based on the user's natural language description of the event, which can improve the efficiency of video tracing. The above-mentioned time information may be the time of the corresponding event in the source video. For example, it may include the event start time and the event end time.

[0057] There may be multiple blockchains, each of which may correspond to a location where a video was taken. For example, taking bank surveillance video as an example, the image acquisition device may be set up in different outlets, and the terminal 102 may build a blockchain based on each outlet, and classify the source video collected by the image acquisition device of the outlet, and store the obtained event location information in the blockchain of the corresponding outlet, thereby realizing the construction of multiple blockchains. When the user needs to trace the source, the blockchain of the outlet where the event occurred can be queried, so as to quickly obtain the corresponding event location information.

[0058] Step S206: query the video database according to the target event location information to obtain the video clip corresponding to the target event.

[0059] The target event location information may be the location information of the video clip where the event that the user needs to trace the source is located. The target event location information includes the event description information and time information of the target event. A plurality of video clips are stored in the video database, and each video clip represents an event. The terminal 102 may query the video database according to the target event location information to obtain the video clip corresponding to the target event. The terminal 102 may display the above video clips, and the terminal 102 may also display the event description information corresponding to the event contained in the video clip and the occurrence time of the event in the video clip, thereby realizing the tracing of the video based on the event description and time.

[0060] In the above-mentioned video tracing method, when receiving a tracing request, the blockchain is queried according to the target event description information and target time information carried in the tracing request to obtain the corresponding target event location information in the blockchain, wherein the blockchain contains multiple blocks for storing different event location information corresponding to different video clips obtained by classifying the source video, and the event location information corresponding to the video clip includes the event description information obtained by describing the set behavior in the video clip and the time information of the above-mentioned video clip in the video. Thus, the video database is queried according to the target event location information to obtain the video clip corresponding to the target event. Compared with the traditional method of manual retrieval, this solution uses the event description information and time information to query each block in the blockchain to obtain the event location information corresponding to the event description information, so that the video clip of the event to be queried can be obtained based on the event location information, thereby improving the efficiency of video tracing.

[0061] In one embodiment, it also includes: obtaining a source video, classifying the source video as an event, obtaining multiple events to be stored and multiple time information to be stored, and constructing multiple blocks in the blockchain according to the multiple time information to be stored; obtaining target events to be stored whose event contents do not overlap from the multiple events to be stored, and generating event description information corresponding to the target events to be stored; obtaining event location information according to the event description information and the time information to be stored, storing the event location information in the block corresponding to the time information to be stored, and obtaining a blockchain.

[0062] In this embodiment, the source video may be a complete video captured by the above-mentioned image acquisition device. The terminal 102 may obtain the source video and classify the source video into events. The classification may be a push-based video stage positioning method. The terminal 102 may classify the source video into events to obtain multiple events to be stored and multiple time information to be stored. The events to be stored may be events included in the source video, and the time information to be stored may be the time occupied by the events in the source video. The above-mentioned time information to be stored may include multiple events, and the terminal 102 may construct multiple blocks in the blockchain based on the multiple time information to be stored.

[0063] After the terminal 102 classifies and obtains multiple events to be stored, it can obtain target events to be stored whose event contents do not overlap from the multiple events to be stored, and generate event description information corresponding to the target events to be stored. The event description information can be a natural language information, so that the terminal 102 can obtain the event location information corresponding to the event based on the above event description information and the time information to be stored. The terminal 102 can obtain multiple event location information based on multiple events to be stored and multiple times to be stored. And the terminal 102 can store each event location information in the block corresponding to its time to be stored. When the terminal 102 stores all event location information in the block corresponding to its time information, the above blockchain can be obtained. Among them, there can be multiple blockchains, and the number of blockchains can be determined based on the collection location corresponding to the source video. For example, taking bank surveillance video as an example, the cameras of bank surveillance video are usually distributed in different outlets, and each outlet can collect the corresponding source video. Then the terminal 102 can construct different blockchains based on the different locations of the outlets to obtain multiple blockchains. When the user needs to trace the source, the event location information can be queried in the blockchain corresponding to the location where the event occurred. The event classification may be performed by a preset identifier, and the event description information may be generated by an LSTM (Long short-term memory) algorithm.

[0064] Through this embodiment, the terminal 102 can obtain multiple events and their times by classifying the source videos, and form event location information based on the description information and time of these events, and store it in the blockchain. Therefore, when the user traces the source, he can query the corresponding video clip only based on the description of the event, thereby improving the efficiency of video tracing.

[0065] In one embodiment, a source video is subjected to event classification to obtain a plurality of events to be stored and a plurality of time information to be stored, including: obtaining a candidate video segment containing a set behavior in the source video according to a classified behavior identifier, obtaining an extended video segment of the event corresponding to the set behavior according to an integrity identifier, obtaining a video segment of the event to be detected and its corresponding timestamp based on the candidate video segment and the extended video segment; aligning the video segment and timestamp of the event to be detected; obtaining a behavior probability and a completion probability corresponding to the video segment of the event to be detected according to the aligned video segment and timestamp of the event to be detected; the behavior probability indicates whether the aligned video segment of the event to be detected is the event to be stored, and the completion probability indicates whether the event to be stored in the aligned video segment of the event to be detected is complete; determining the event to be stored and the time information to be stored containing the complete event to be stored according to the behavior probability and the completion probability.

[0066] In this embodiment, the terminal 102 can obtain video clips corresponding to multiple events by performing push-type classification on the source video. The terminal 102 can obtain the candidate video clips containing the set behavior in the source video through the classification behavior identifier, and obtain the extended video clip of the event corresponding to the above-mentioned set behavior according to the integrity identifier, so that the terminal 102 can obtain the video clip of the event to be detected and its corresponding timestamp in the source video based on the candidate video clip and the extended video clip. For example, taking a bank surveillance video as an example, a user withdrawing money from an ATM can be regarded as an event, and in the video clip of the event, the action of the user withdrawing money can be regarded as a candidate video clip in the video. The terminal 102 can obtain the extended video clip of the above-mentioned withdrawal event based on the user's withdrawal action through the integrity identifier, such as the clip of the user walking into and leaving the ATM. Then, the terminal 102 can obtain the video clip of the user withdrawing money event and its timestamp in the source video based on the above-mentioned candidate clip and the extended clip. Specifically, if Figure 3 As shown, Figure 3 FIG. 1 is a flow chart of an event classification step in an embodiment. When performing stage positioning on the source video, the terminal 102 can classify the video into candidate areas and extended areas through a classification behavior identifier and an integrity identifier, and perform sampling of sparse segments, so that the terminal 102 can use the sparse segments as the above-mentioned event video segments to be detected.

[0067] The terminal 102 may also align the above-mentioned video clip of the event to be detected and the timestamp. For example, in one embodiment, aligning the video clip of the event to be detected and the timestamp includes: performing pyramid pooling processing on the timestamp, inputting the processed timestamp and the video clip of the event to be detected into a preset recurrent neural network, and aligning the video clip of the event to be detected and the timestamp through the preset recurrent neural network. In this embodiment, during the alignment process, the terminal 102 may perform pyramid pooling processing on the timestamp, and input the processed timestamp and the video clip of the event to be detected into a preset RNN (Recurrent Neural Network) network, thereby aligning the video clip of the event to be detected and the timestamp through the preset RNN network. Specifically, as Figure 3 As shown, the terminal 102 can perform pyramid pooling based on the timestamps of the events in the above-mentioned various stages, and align the video time structure and generate the text structure through RNN network training.

[0068] After alignment, the terminal 102 can also obtain the behavior probability and completion probability corresponding to the video clip of the event to be detected based on the aligned video clip of the event to be detected and the timestamp. The behavior probability indicates whether the aligned video clip of the event to be detected is the event to be stored, such as whether it is the action of the user withdrawing money; the completion probability indicates whether the event to be stored in the aligned video clip of the event to be detected is complete, such as whether it includes the whole process from the user appearing on the screen to the user leaving the screen. Specifically, Figure 3 As shown, after aligning the time, the terminal 102 can calculate the key frame according to the precise time positioning, and obtain the behavior probability and the completion degree probability by processing the regional features through the classifier. The terminal 102 determines whether the timestamp is the key frame time through the above two probabilities. For example, when the above two probabilities are greater than or equal to the preset probability threshold, the above timestamp is determined to be the key frame time. After the terminal 102 determines the key frame, the timestamp of the key frame can be retrieved corresponding to the relevant block height to obtain the corresponding block, and the key frame probability is recorded in the blockchain. Therefore, the terminal 102 can determine multiple events to be stored and the time information to be stored containing the complete events to be stored from the source video according to the above behavior probability and completion degree probability.

[0069] Through this embodiment, the terminal 102 can classify events of the source video based on a preset identifier, and determine the complete event and the complete time information to be stored through the behavior probability and the completion degree probability, and trace the source based on the complete event and the time information to be stored, which can improve the efficiency of video tracing.

[0070] In one embodiment, multiple blocks in a blockchain are constructed according to multiple time information to be stored, including: for each time information to be stored, determining the duration corresponding to the time information to be stored according to the timestamp contained in the time information to be stored, and generating a block of a size corresponding to the duration in the blockchain to be constructed; and obtaining a blockchain according to the multiple blocks.

[0071] In this embodiment, the terminal 102 can construct multiple blocks to form the above-mentioned blockchain based on the above-mentioned multiple time information to be stored. The above-mentioned events to be stored may include multiple, and the corresponding time information to be stored may also have multiple. For each time information to be stored, the terminal 102 can determine the duration corresponding to the time information to be stored according to the timestamp contained in the time information to be stored. For example, the terminal 102 can obtain the start time and end time of the event to be stored contained in the time information to be stored in the video, thereby obtaining multiple timestamps of the event to be stored, and the terminal 102 determines the duration of the time information to be stored based on the multiple timestamps. After the terminal 102 determines the duration, it can determine the size of the block of the event to be stored corresponding to the time information to be stored in the blockchain to be constructed based on the duration. Since there are multiple events to be stored, the terminal 102 can construct multiple blocks through the above method, so that the terminal 102 can obtain the above-mentioned blockchain according to multiple blocks. Specifically, the terminal 102 can divide the time according to the sampled fragments of the source video, form the timestamp in the blockchain according to the divided time, construct the block in the blockchain based on the timestamp, and record the block content according to the timestamp mark block height. In addition, the terminal 102 may also record the probability of whether the image frame in the video corresponding to the above-mentioned judgment timestamp is a key frame in the blockchain. For example, the terminal 102 may judge whether the image frame corresponding to the timestamp is a key frame based on the behavior probability and the completion degree probability. If so, the timestamp of the key frame is searched for the relevant block height to find the block, and the key frame probability is recorded in the blockchain.

[0072] Through this embodiment, the terminal 102 can determine the duration in the block based on multiple timestamps in the time information to be stored, so as to construct a block corresponding to the time information to be stored, thereby improving the query efficiency during video tracing.

[0073] In one embodiment, a target event to be stored whose event content does not overlap is obtained from a plurality of events to be stored, including: for each frame of video image in a video clip corresponding to the event to be stored, obtaining an image frame to be detected whose similarity with other frames of video image in the video clip is less than a preset similarity threshold; inputting the image frame to be detected into a preset probability function to obtain a corresponding non-overlapping probability value, and determining the target image frame to be detected whose event content does not overlap according to a comparison result of the non-overlapping probability value and a preset hyperparameter; and obtaining a target event to be stored according to a plurality of target image frames to be detected.

[0074] In this embodiment, when the terminal 102 performs push-type classification on the events in the source video, there may be overlapping images, so the terminal 102 needs to deduplicate the events to be stored to obtain target events to be stored whose event contents do not overlap. The video clip corresponding to the above-mentioned event to be stored may include multiple frames of images. For each frame of video image in the video clip of the event to be stored, the terminal 102 can obtain the image frame to be detected whose similarity with other frames of video images in the video clip is less than a preset similarity threshold. The terminal 102 can input the image frame to be detected into a preset probability function to obtain the corresponding non-overlapping probability value characterizing that the image frame to be detected is a non-overlapping image. The terminal 102 can obtain the comparison result of the non-overlapping probability value and the preset hyperparameter, and determine whether the image frame is a target image frame to be detected whose event contents do not overlap according to the comparison result. Wherein, there can be multiple video image frames in the above-mentioned video clip, and the terminal 102 can determine whether each video image frame is a target image frame to be detected, so that the terminal 102 can obtain the target event to be stored based on multiple target image frames to be detected.

[0075] Among them, the above-mentioned obtained video image frame and the image frame to be detected whose similarity with other frames of video images in the video clip is less than the preset similarity threshold can be obtained by LSTM. For example, in one embodiment, the image frame to be detected whose similarity with other frames of video images in the video clip is less than the preset similarity threshold is obtained, including: if the frame of video image is the first frame in the video clip, according to the visual features of the frame of video image, the time information occupied by the frame of image in the video clip and the generated description information, the image frame to be detected is obtained by the long short-term memory algorithm; if the frame of video image is not the first frame in the video clip, according to the previous frame of image frame to be detected corresponding to the frame of video image, the visual features of the frame of video image, the time information occupied by the frame of image in the video clip and the generated description information, the image frame to be detected is obtained by the long short-term memory algorithm.

[0076] In this embodiment, when the terminal 102 obtains the image frame to be detected from the video clip, different calculation methods can be used based on whether the frame of video image is the first frame. If the terminal 102 detects that the frame of video image is the first frame in the video clip, the terminal 102 can obtain the image frame to be detected through the above-mentioned LSTM algorithm based on the visual features of the frame of video image, the time information occupied by the frame of image in the above-mentioned video clip, and the generated description information. If the terminal 102 detects that the frame of image is not the first frame in the video clip, the terminal 102 can obtain the image frame to be detected through the above-mentioned LSTM algorithm based on the previous frame of image frame to be detected corresponding to the frame of video image, the visual features of the frame of video image, the time information occupied by the frame of video image in the video clip, and the generated description information.

[0077] Specifically, the terminal 102 can generate descriptive words for each frame of the video image in units of words, that is, the terminal 102 can select events and form text, and generate the next word based on the video time structure and the word sequence that has been generated to ensure the coherence between the generated sentences. The terminal 102 can use LSTM to select segments with independent events that do not overlap with previous events to avoid a large amount of redundancy in the generated description. Among them, the image frame to be detected obtained by the terminal 102 through LSTM can be recorded as a vector: Among them, h t-1 is the previous image frame to be detected, if h t is the first frame, the above vector may not contain h t-1 , the above input v t is the visual feature extracted by the video frame through the video stage positioning method, r t is a distance feature, similar to an image mask, represented by a binary mask that represents the normalized time span of the segment relative to the entire duration. kt is the feature of the sentence that has been output, and is the last potential state of the subtitle generation network that generates the previous sentence, so that the terminal 102 can take into account the words that have been spoken when selecting events, and t is the time at this time. The terminal 102 obtains the above-mentioned image frame to be detected h t After that, the non-overlapping probability value can be obtained by the following formula: Among them, p t The probability that the fragment contains relevant and unique information, that is, the above-mentioned non-overlapping probability value, is a hyperparameter. When the above-mentioned non-overlapping probability value is greater than the entire threshold, for example, greater than the above-mentioned hyperparameter, the terminal 102 determines that the fragment is a fragment that needs to generate a description, that is, as a target event to be stored.

[0078] Through the above embodiment, the terminal 102 can determine the non-overlapping target events to be stored in the video image of the event to be detected by LSTM, thereby improving the query efficiency during video tracing.

[0079] In one embodiment, event description information corresponding to a target event to be stored is generated, including: obtaining multiple sub-events contained in the target event to be stored, and for each sub-event, determining a sub-event description corresponding to the sub-event by a long short-term memory algorithm based on visual features of a video clip corresponding to the sub-event and a generated sub-event description; and obtaining event description information corresponding to the target event to be stored based on the multiple sub-event descriptions.

[0080] In this embodiment, the terminal 102 can generate an event description for each non-overlapping target event to be stored. The above-mentioned target event to be stored may include multiple sub-events, for example, each action of the user can constitute a sub-event. When generating an event description, the terminal 102 can obtain multiple sub-events contained in the target event to be stored. For each sub-event, the terminal 102 can determine the sub-event description corresponding to the sub-event through the LSTM algorithm based on the visual features of the video clip corresponding to the sub-event and the generated sub-event description. Since there can be multiple sub-events, the terminal 102 can obtain the event description information corresponding to the target event to be stored based on the sub-event descriptions of multiple sub-events. Specifically, after the terminal 102 obtains a segment with an independent event through LSTM, a single-sentence description of the segment event can be obtained based on LSTM. The terminal 102 can splice all the sentences output by LSTM into a long paragraph to represent the description of the event content in the entire long video. The specific calculation formula can be as follows: in It means the lth step when describing the kth event, k can be the number of sub-events, Visual features representing a sub-region of an event. Represents the word generated in the previous step.

[0081] Through this embodiment, the terminal 102 can generate push-type event description information for the target event to be stored through the LSTM algorithm, and generate the description information of the entire target event to be stored by splicing single-sentence descriptions, so that when tracing the source of the video, it can be traced based on the description information, thereby improving the query efficiency of video tracing.

[0082] In one embodiment, event location information is obtained based on event description information and time information to be stored, and the event location information is stored in a block corresponding to the time information to be stored to obtain a blockchain, including: concatenating the event description information and the time information to be stored to obtain corresponding coordinates as event location information, and storing the event location information in the form of coordinates in a block corresponding to the time information to be stored to obtain a blockchain.

[0083] In this embodiment, after obtaining the above-mentioned event description information and the time information to be stored, the terminal 102 can splice the event description information and the time information to be stored to obtain the corresponding coordinates as the above-mentioned event location information. The terminal 102 can store the event location information in the block corresponding to the time to be stored in the form of the above-mentioned coordinates. Since the above-mentioned events to be stored may include multiple events, the event location information generated by the terminal 102 can also be multiple. The terminal 102 can store each event location information in the block corresponding to its time information to be stored, thereby obtaining the corresponding blockchain. Specifically, the text description record w, timestamp t and other information generated by the terminal 102 can form the coordinates L(w, t), which are recorded through the blockchain. The terminal 102 can also analyze the text and store these text information and timestamp information as identity authentication information. When the user encounters the same situation again, it can be traced back through the blockchain.

[0084] Through this embodiment, the terminal 102 can store the events to be stored and the time information to be stored in the form of coordinates, so that the terminal 102 can query the corresponding event location information through the description information of the event when tracing the source of the video, thereby improving the query efficiency of tracing the source of the video.

[0085] In one embodiment, Figure 4 As shown, Figure 4 It is a flow chart of the video tracing method in another embodiment. Taking the video tracing method applied to bank surveillance video as an example, it includes the following process: as a kind of unstructured data, bank surveillance video is difficult to structure the content and features of the video image. Surveillance cameras are deployed in various places related to banks, such as outlets, data centers, vaults, etc., and real-time work surveillance video. Terminal 102 can convert the video content, realize the text conversion of the video content based on the blockchain timestamp and the fusion of LSTM and RNN, and fuse the converted text with the timestamp to form coordinates, which are stored on the blockchain to realize tracing. Specifically, it can be a push-type event classification process, including a push-type fusion blockchain timestamp based on the video stage to generate a word sequence method and a work safety tracing system based on the blockchain for banks. In the actual construction of the blockchain, the terminal 102 can obtain each surveillance video in the bank through the camera, push-type based on the video stage to locate the event, obtain the time and key frame positioning and interact with the blockchain, select the independent events in the video based on LSTM, and then splice the text with independent event information.

[0086] Specifically, the terminal 102 can perform push-based video stage positioning events on each surveillance video in the bank, align the video time structure through the RNN network and generate a text structure. The terminal 102 can obtain the timestamp of the event by selecting key frames and record the timestamp of the key frames in the blockchain. The terminal 102 can also select video clips corresponding to non-overlapping independent events through LSTM, and obtain single-sentence descriptions of these clips through LSTM, so that the terminal 102 can combine the generated text description and timestamp information into coordinates and record them in the blockchain. Users can trace the source of the event by inputting the description information of the event.

[0087] In addition, in one embodiment, an application embodiment is provided, such as Figure 5 As shown, Figure 5 It is a flow chart of the video tracing method in another embodiment. Taking the user withdrawing money from the outlet as an example, the interactive process between the terminal 102 and the user includes: the user uses the outlet to withdraw money service, and from the time when the user enters the video screen, the terminal 102 locates and records the action information, and uses the push-type fusion blockchain timestamp to generate a word sequence method based on the video stage to convert it into text, and then the time is matched, and the content of the withdrawal process is traced on the blockchain. Since each window or ATM machine can only allow one person to operate, it can correspond to the user account operation record. The specific steps are: the user withdraws money from the bank account, the blockchain records the time, location and operation of the withdrawal, and the node of the outlet records this process in a block. The terminal 102 of the outlet compares the operation content according to the time on the same day, and displays the operation content of this time to the customer. The user can trace the video recorded by the outlet through the above-mentioned video tracing method. In addition, the terminal 102 can also maintain the bank data center and record the maintenance process. For example, for the daily maintenance and cleaning of the bank data center, as well as the replacement of hardware facilities and software operation, from the time the staff enters the video screen to the time they leave the screen, the terminal 102 can convert the video source traceability method into text, store the key information according to the timestamp, and compare the text according to the daily reporting requirements, and verify the source based on the blockchain timestamp to ensure safety. In addition, the terminal 102 can also store and trace the video in the bank vault and other places. For example, the terminal 102 can connect the bank vault's warehouse, storage room, bill processing site, in and out warehouse handover site, main channel and other monitoring installations to this system and the alarm center control system. The terminal 102 can store the text formed in the push-based video stage through the blockchain node every day to ensure that different time periods can be safely stored. At the same time, the terminal 102 retrieves the traceability content through the coordinates in combination with the text storage time.

[0088] Through the above embodiment, the terminal 102 obtains event location information corresponding to the event description information by querying each block in the blockchain using the event description information and time information, so as to obtain the video clip of the event to be queried based on the event location information, thereby improving the efficiency of video tracing.

[0089] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0090] Based on the same inventive concept, the embodiment of the present application also provides a video tracing device for implementing the video tracing method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more video tracing device embodiments provided below can refer to the limitations of the video tracing method above, and will not be repeated here.

[0091] In one embodiment, Figure 6 As shown, a video source tracing device is provided, including: a receiving module 500, a source tracing module 502 and a query module 504, wherein:

[0092] The receiving module 500 is used to receive a tracing request for a target event, and obtain the target event description information and target time information carried in the tracing request.

[0093] The traceability module 502 is used to query the blockchain according to the target event description information and the target time information to obtain the corresponding target event location information; the blockchain contains multiple blocks for storing different event location information corresponding to different video clips; the event location information corresponding to the video clip includes the event description information of the video clip and its time information; the video clip is obtained based on event classification of the source video, the event description information represents the description of the set behavior in the video clip, and the time information represents the time of the video clip in the video.

[0094] The query module 504 is used to query the video database according to the target event location information to obtain the video clip corresponding to the target event.

[0095] In one embodiment, the above-mentioned device also includes: a construction module, which is used to obtain a source video, classify the source video into events, obtain multiple events to be stored and multiple time information to be stored, and construct multiple blocks in the blockchain according to the multiple time information to be stored; obtain target events to be stored whose event contents do not overlap from the multiple events to be stored, and generate event description information corresponding to the target events to be stored; obtain event location information according to the event description information and the time information to be stored, store the event location information in the block corresponding to the time information to be stored, and obtain a blockchain.

[0096] In one embodiment, the above-mentioned construction module is specifically used to obtain a candidate video segment containing a set behavior in the source video according to a classified behavior identifier, obtain an extended video segment of an event corresponding to the set behavior according to an integrity identifier, and obtain a video segment of an event to be detected and its corresponding timestamp based on the candidate video segment and the extended video segment; align the video segment and timestamp of the event to be detected; obtain the behavior probability and completion probability corresponding to the video segment of the event to be detected according to the aligned video segment and timestamp of the event to be detected; the behavior probability indicates whether the aligned video segment of the event to be detected is the event to be stored, and the completion probability indicates whether the event to be stored in the aligned video segment of the event to be detected is complete; determine the event to be stored and the time information to be stored containing the complete event to be stored according to the behavior probability and the completion probability.

[0097] In one embodiment, the above-mentioned construction module is specifically used to perform pyramid pooling processing on the timestamp, input the processed timestamp and the video clip of the event to be detected into a preset recurrent neural network, and align the video clip of the event to be detected and the timestamp through the preset recurrent neural network.

[0098] In one embodiment, the above-mentioned construction module is specifically used to determine, for each time information to be stored, the duration corresponding to the time information to be stored according to the timestamp contained in the time information to be stored, and generate a block of a size corresponding to the duration in the blockchain to be constructed; and obtain a blockchain based on multiple blocks.

[0099] In one embodiment, the above-mentioned construction module is specifically used to obtain, for each frame of video image in the video clip corresponding to the event to be stored, an image frame to be detected whose similarity with other frames of video image in the video clip is less than a preset similarity threshold; input the image frame to be detected into a preset probability function to obtain a corresponding non-overlapping probability value, and determine the target image frame to be detected whose corresponding event content does not overlap based on the comparison result of the non-overlapping probability value and the preset hyperparameter; and obtain the target event to be stored based on multiple target image frames to be detected.

[0100] In one embodiment, the above-mentioned construction module is specifically used to obtain the image frame to be detected by using a long short-term memory algorithm if the frame of video image is the first frame in the video clip, based on the visual features of the frame of video image, the time information occupied by the frame of image in the video clip and the generated description information; if the frame of video image is not the first frame in the video clip, obtain the image frame to be detected by using a long short-term memory algorithm based on the previous frame of image frame to be detected corresponding to the frame of video image, the visual features of the frame of video image, the time information occupied by the frame of image in the video clip and the generated description information.

[0101] In one embodiment, the above-mentioned construction module is specifically used to obtain multiple sub-events contained in the target event to be stored, and for each sub-event, determine the sub-event description corresponding to the sub-event through a long short-term memory algorithm based on the visual features of the video clip corresponding to the sub-event and the generated sub-event description; based on the multiple sub-event descriptions, obtain event description information corresponding to the target event to be stored.

[0102] In one embodiment, the above-mentioned construction module is specifically used to splice the event description information and the time information to be stored to obtain the corresponding coordinates as the event location information, and store the event location information in the form of coordinates in the block corresponding to the time information to be stored to obtain a blockchain.

[0103] Each module in the above-mentioned video source tracing device can be implemented in whole or in part by software, hardware and their combination. Each of the above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0104] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 7As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a video tracing method is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a button, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse.

[0105] Those skilled in the art will understand that Figure 7 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0106] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the above-mentioned video tracing method when executing the computer program.

[0107] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned video tracing method is implemented.

[0108] In one embodiment, a computer program product is provided, including a computer program, which implements the above-mentioned video tracing method when executed by a processor.

[0109] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0110] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0111] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0112] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A video tracing method, characterized in that: The method comprises: Receiving a tracing request for a target event, obtaining target event description information and target time information carried in the tracing request; According to the target event description information and the target time information, a blockchain is queried to obtain the corresponding target event location information; the blockchain contains a plurality of blocks for storing different event location information corresponding to different video clips; the event location information corresponding to the video clip includes the event description information of the video clip and its time information; the video clip is obtained based on event classification of the source video, and the event classification of the source video includes the steps of obtaining a candidate video clip containing a set behavior in the source video according to a classification behavior identifier, and obtaining an extended video clip of the event corresponding to the set behavior based on an action through an integrity identifier, and obtaining a video clip of the event to be detected and its corresponding timestamp based on the candidate video clip and the extended video clip; the event description information represents a description of the set behavior in the video clip, and the time information represents the time of the video clip in the video; The video database is queried according to the target event location information to obtain a video clip corresponding to the target event.

2. The method according to claim 1, characterized in that The method further comprises: Acquire a source video, classify the source video into events, obtain multiple events to be stored and multiple time information to be stored, and construct multiple blocks in the blockchain according to the multiple time information to be stored; Acquire a target event to be stored whose event contents do not overlap from the multiple events to be stored, and generate event description information corresponding to the target event to be stored; Event location information is obtained according to the event description information and the time information to be stored, and the event location information is stored in a block corresponding to the time information to be stored to obtain the blockchain.

3. The method according to claim 2, characterized in that The event classification of the source video to obtain a plurality of events to be stored and a plurality of time information to be stored includes: Obtain a candidate video segment containing a set behavior in the source video according to a classification behavior identifier, obtain an extended video segment of an event corresponding to the set behavior according to an integrity identifier, and obtain a video segment of an event to be detected and its corresponding timestamp based on the candidate video segment and the extended video segment; Aligning the video clip of the event to be detected and the timestamp; According to the aligned video clips of the event to be detected and the timestamps, the behavior probability and the completion degree probability corresponding to the video clips of the event to be detected are obtained; the behavior probability indicates whether the aligned video clips of the event to be detected are the events to be stored, and the completion degree probability indicates whether the events to be stored in the aligned video clips of the event to be detected are complete; According to the behavior probability and the completion degree probability, the events to be stored and the time information to be stored including the complete events to be stored are determined.

4. The method according to claim 3, characterized in that The aligning the video clip of the event to be detected and the timestamp includes: The timestamp is subjected to pyramid pooling processing, the processed timestamp and the video clip of the event to be detected are input into a preset recurrent neural network, and the video clip of the event to be detected and the timestamp are aligned through the preset recurrent neural network.

5. The method according to claim 3, characterized in that: The step of constructing a plurality of blocks in the blockchain according to the plurality of time information to be stored includes: For each time information to be stored, determine the duration corresponding to the time information to be stored according to the timestamp contained in the time information to be stored, and generate a block of a size corresponding to the duration in the blockchain to be constructed; The blockchain is obtained based on the multiple blocks.

6. The method according to claim 2, characterized in that The step of acquiring a target event to be stored whose event contents do not overlap from the multiple events to be stored includes: For each frame of video image in the video clip corresponding to the event to be stored, obtaining an image frame to be detected whose similarity with other frames of video image in the video clip is less than a preset similarity threshold; Input the image frame to be detected into a preset probability function to obtain a corresponding non-overlapping probability value, and determine the target image frame to be detected whose corresponding event content does not overlap according to the comparison result of the non-overlapping probability value and the preset hyperparameter; According to the plurality of target to-be-detected image frames, a target to-be-stored event is obtained.

7. The method according to claim 6, characterized in that: The step of obtaining the image frame to be detected whose similarity with other frame video images in the video segment is less than a preset similarity threshold comprises: If the frame of video image is the first frame in the video clip, according to the visual features of the frame of video image, the time information occupied by the frame of video image in the video clip and the generated description information, the image frame to be detected is obtained by using a long short-term memory algorithm; If the video image frame is not the first frame in the video clip, the image frame to be detected is obtained by a long short-term memory algorithm based on the previous image frame to be detected corresponding to the video image frame, the visual features of the video image frame, the time information occupied by the video image frame in the video clip and the generated description information.

8. The method according to claim 2, characterized in that: The generating event description information corresponding to the target event to be stored includes: Acquire multiple sub-events included in the target event to be stored, and for each of the sub-events, determine the sub-event description corresponding to the sub-event by using a long short-term memory algorithm according to the visual features of the video clip corresponding to the sub-event and the generated sub-event description; According to the multiple sub-event descriptions, event description information corresponding to the target event to be stored is obtained.

9. The method according to claim 2, characterized in that: The obtaining of event location information according to the event description information and the time information to be stored, storing the event location information in a block corresponding to the time information to be stored, and obtaining the blockchain includes: The event description information and the time information to be stored are concatenated to obtain corresponding coordinates as event location information, and the event location information is stored in the block corresponding to the time information to be stored in the form of the coordinates to obtain the blockchain.

10. A video source tracing device, characterized in that: The device comprises: A receiving module, configured to receive a tracing request for a target event, and obtain target event description information and target time information carried in the tracing request; A traceability module is used to query the blockchain according to the target event description information and the target time information to obtain the corresponding target event location information; the blockchain contains multiple blocks for storing different event location information corresponding to different video clips; the event location information corresponding to the video clip includes the event description information of the video clip and its time information; the video clip is obtained based on event classification of the source video, and the event classification of the source video includes obtaining a candidate video clip containing a set behavior in the source video according to a classification behavior identifier, and obtaining an extended video clip of the event corresponding to the set behavior based on an action through an integrity identifier, and obtaining a video clip of the event to be detected and its corresponding timestamp based on the candidate video clip and the extended video clip; the event description information represents the description of the set behavior in the video clip, and the time information represents the time of the video clip in the video; The query module is used to query the video database according to the target event location information to obtain the video clip corresponding to the target event.

11. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 9 are implemented.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 9 are implemented.