Field electric power supervision service behavior tracing method and system based on block chain

Abstract

The invention belongs to the technical field related to electric power, and provides a method and a system for tracing a field electric power supervision service behavior based on a block chain in order to solve the problem of difficulty in video tracing in the current electric power service, the video data of an electric power monitoring area is stored in the block chain, abnormal events are extracted from the collected video data of the electric power monitoring area, and the abnormal events are stored in the block chain. And matching the extracted abnormal event with the complete video data stored in the power monitoring area of the block chain, and calling the video data corresponding to the abnormal event on the block chain according to a matching result. When the data is stored in the block chain server and confirmed, the data cannot be tampered or deleted, so that the integrity and security of the data can be ensured, the data is prevented from being maliciously tampered or damaged, and the tracing time can be prolonged; and the abnormal event in the video is extracted and the corresponding video data is called, so that more comprehensive and three-dimensional event restoration is provided.

Description

Technical Field

[0001] The present invention belongs to the technical field related to electricity, and in particular to a tracing method and system for on-site electricity supervision service behavior based on blockchain. Background Art

[0002] The statements herein merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] Power services involve the operation and management of power systems. When performing on-site power services, there may be safety hazards and risks. Video can be used to monitor on-site power services and understand the course of accidents in a timely manner, thereby ensuring the safety of personnel and equipment.

[0004] When existing power services use video surveillance, video data is usually stored in full, which also includes video data from some unimportant time periods. When the memory reaches the upper limit, new video data will overwrite the old video data, so the scope of video tracing is limited; When providing electricity services, an abnormal event is usually judged as an abnormal event and video capture is started only after it occurs. This may result in missing video time of the abnormal event, making it impossible to understand the beginning and end of the abnormal event more clearly. At the same time, it is also impossible to automatically match and extend the video time based on the captured video clips of the abnormal event, making the tracing of the clip video data more troublesome. Summary of the invention

[0005] In order to overcome the shortcomings of the above-mentioned prior art, the present invention provides a blockchain-based on-site power supervision service behavior tracing method and system, which saves the video to the blockchain, extracts abnormal events in the video, and retrieves the corresponding video data based on the extracted abnormal events, thereby providing a more comprehensive and three-dimensional event restoration.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: In a first aspect, the present invention provides a blockchain-based on-site power supervision service behavior tracing method, comprising: S1. Collect video data of the power monitoring area and save it to the blockchain; S2. Extracting abnormal events from the collected video data of the power monitoring area; wherein the abnormal events include abnormal personnel identity, abnormal behavior and abnormal working environment status; S3, matching the extracted abnormal events with the complete video data of the power monitoring area saved to the blockchain; S4. Retrieve the video data corresponding to the abnormal event on the blockchain according to the matching result.

[0007] Furthermore, the abnormal events are extracted from the collected video data of the power monitoring area; wherein the abnormal events include abnormal personnel identity, abnormal behavior and abnormal working environment state, and the specific steps are: S21, performing personnel identification on the collected video data of the power monitoring area; S22, identifying abnormal behavior on the collected video data of the power monitoring area; S23: Identify abnormal working environment status based on the collected video data of the power monitoring area.

[0008] Furthermore, the specific steps of performing personnel identification on the collected video data of the power monitoring area are as follows: S211, extracting body shape features, external work clothes features, voiceprint features and hair features of the people in the video; S212, performing weighted fusion on the extracted features to obtain fusion features of the people in the video; S213: Input the fusion features of the persons in the video into the trained classifier to obtain the identity recognition results of the persons in the video.

[0009] Furthermore, the specific steps of performing abnormal behavior recognition on the collected video data of the power monitoring area are as follows: S221, extracting image feature representation in the video through a convolutional neural network; S222, using a target detection algorithm to identify whether a person in the video image is wearing safety equipment through bounding box regression and a classifier; S223, determining whether the person is in the specified working area according to the relationship between the position of the person in the video and the specified working area; S224, using an anomaly detection algorithm to detect the movement of people in the video, and determine whether there is abnormal movement of people; S225. If it is detected that at least one of the following situations occurs: the person is not wearing safety equipment, the person is not in the specified working area, or the person moves abnormally, it is determined to be abnormal behavior.

[0010] Furthermore, the relationship between the position of the person in the video and the specified working area is used to determine whether the person is in the specified working area. The specific formula is: , Among them, T is the number of intersection points, P is the point of the staff's current position, P is the boundary line segment of the specified area, intersect represents a function for determining whether a point intersects with a boundary line segment, and it means summing up all boundary line segments to obtain the total number of intersections between all boundary line segments and point P; when the number of intersection points is an even number, it means that the staff is outside the specified area. When the staff is outside the specified area, it is determined to be working in an abnormal area; when the number of intersection points is an odd number, it means that the staff is working in the specified area.

[0011] Furthermore, the use of an abnormality detection algorithm to detect the movement of people in the video to determine whether there is abnormal movement of people is specifically as follows: S2241, obtaining a position change of the staff member by calculating the Euclidean distance according to the position of the staff member in the current frame and the position of the staff member in the previous frame; S2242. Determine whether the worker has moved abnormally based on the worker's position change and the set threshold range of normal operation.

[0012] Furthermore, the specific steps of identifying the abnormal working environment state of the collected video data of the power monitoring area are as follows: S231. Detect abnormal sounds in the environment through sound recognition technology; wherein the abnormal sounds include prolonged abnormal sounds from power facilities and loud sounds in the environment.

[0013] Furthermore, the extracted abnormal events are matched with the complete video data of the power monitoring area saved in the blockchain, and the specific steps are as follows: S31, for the video of the identified abnormal environment, matching the matched complete video with the complete video around the work scene in the same time period, and according to the start and end timestamps of the video segment A, extracting the sub-segment of the same time period as the video segment A from other complete videos around the work scene to generate the video segment B, and combining the video segment A and the video segment B into a video segment to form a multi-view video set; S32. For videos showing abnormal identification of personnel and when personnel are not working in the standard area, an index is created for each identified object based on the identified facial information, and all related videos showing the target person in video segment A appearing within a set time window are searched in the complete video data based on the created object index. The video capture time is set, video segments are generated, and videos showing abnormal identification of personnel are compiled into a video set to form a video set showing abnormal identification of personnel.

[0014] In a second aspect, the present invention provides a blockchain-based on-site power supervision service behavior tracing system, including: Video acquisition module, used to collect video data of the power monitoring area and save it to the blockchain; An abnormal event extraction module is used to extract abnormal events from the collected video data of the power monitoring area; wherein the abnormal events include abnormal personnel identity, abnormal behavior and abnormal working environment status; An event matching module, used to match the extracted abnormal events with the complete video data of the power monitoring area saved to the blockchain; The tracing platform is used to retrieve the video data corresponding to the abnormal event on the blockchain based on the matching results.

[0015] Furthermore, the abnormal event extraction module includes: A personnel identification module is used to identify personnel from the collected video data of the power monitoring area; An abnormal behavior recognition module is used to recognize abnormal behaviors of the collected video data of the power monitoring area; The abnormal working environment status module is used to identify the abnormal working environment status of the collected video data of the power monitoring area.

[0016] Furthermore, the event matching module includes: Construct a multi-view video set module, which is used to match the matched complete video with the complete video around the work scene in the same time period for the video of the identified abnormal environment, and extract the sub-segment of the same time period as the video segment A from other complete videos around the work scene according to the start and end timestamps of the video segment A to generate the video segment B, and combine the video segment A and the video segment B into a video segment to form a multi-view video set; Construct a video collection module for abnormal personnel identity, which is used for videos of abnormal personnel identity recognition and personnel not working in the standardized area. According to the recognized facial information, an index is established for each identified object, and all related videos appearing in the target person in video segment A within the set time window are searched in the complete video data according to the established object index. The video capture time is set, and video segments are generated. The videos with abnormal personnel are made into a video collection to form a video collection of abnormal personnel identity.

[0017] One or more of the above technical solutions have the following beneficial effects: In the present invention, the video is saved to the blockchain. Since the data on the blockchain is saved in an encrypted and distributed storage manner, once the data is stored on the blockchain server and confirmed, it cannot be tampered with or deleted, thereby ensuring the integrity and security of the data, preventing the data from being maliciously tampered with or damaged, and increasing the traceability time; abnormal events in the video are extracted, and the corresponding video data is retrieved according to the extracted abnormal events, thereby providing a more comprehensive and three-dimensional event restoration.

[0018] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0020] Figure 1 A flowchart of a blockchain-based on-site power supervision service behavior tracing method of the present invention; Figure 2 This is a block diagram of the blockchain-based on-site power supervision service behavior tracing system of the present invention. DETAILED DESCRIPTION

[0021] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0022] It should be noted that the terms used herein are for describing specific embodiments only and are not intended to be limiting of exemplary embodiments according to the present invention.

[0023] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0024] Embodiment 1 like Figure 1 As shown, this embodiment discloses a blockchain-based on-site power supervision service behavior tracing method, including: S1. Collect video data from the power monitoring area and save it to the blockchain.

[0025] S2. Extract abnormal events from the collected video data of the power monitoring area; wherein the abnormal events include abnormal personnel identity, abnormal behavior and abnormal working environment status.

[0026] In this embodiment, abnormal events are extracted from the collected video data of the power monitoring area; wherein the abnormal events include abnormal personnel identity, abnormal behavior and abnormal working environment state. The specific steps are: S21. Performing personnel identification on the collected video data of the power monitoring area.

[0027] In this embodiment, the specific steps of performing personnel identification on the collected video data of the power monitoring area are as follows: S211, extracting body shape features, external work clothes features, voiceprint features and hair features of the people in the video; S212, performing weighted fusion on the extracted features to obtain fusion features of the people in the video; S213: Input the fusion features of the persons in the video into the trained classifier to obtain the identity recognition results of the persons in the video.

[0028] S22: Identify abnormal behaviors on the collected video data of the power monitoring area.

[0029] In this embodiment, the specific steps of performing abnormal behavior recognition on the collected video data of the power monitoring area are as follows: S221, extracting image feature representation in the video through a convolutional neural network; S222, using a target detection algorithm to identify whether a person in the video image is wearing safety equipment through bounding box regression and a classifier; S223, determining whether the person is in the specified working area according to the relationship between the position of the person in the video and the specified working area; In this embodiment, whether the person is in the prescribed working area is determined based on the relationship between the position of the person in the video and the prescribed working area. The specific formula is: ; Among them, T is the number of intersection points, P is the point where the staff member is currently located, is the boundary segment of the specified area, intersect Indicates judgment point Is it consistent with the boundary segment? The intersection function, For all boundary segments The sum is calculated to obtain the total number of intersections between all boundary segments and point P. When the number of intersections is an even number, it means that the staff is outside the specified area. When the staff is outside the specified area, it is determined to be working in an abnormal area. When the number of intersections is an odd number, it means that the staff is working in the specified area. S224: Detect the movement of people in the video using an anomaly detection algorithm to determine whether there is abnormal movement of the people.

[0030] In this embodiment, an abnormality detection algorithm is used to detect the movement of people in the video to determine whether there is abnormal movement of people, specifically: S2241, obtaining a position change of the staff member by calculating the Euclidean distance according to the position of the staff member in the current frame and the position of the staff member in the previous frame; S2242, judging whether the worker moves abnormally according to the position change of the worker and the set threshold range of normal operation; S225. If it is detected that at least one of the following situations occurs: the person is not wearing safety equipment, the person is not in the specified working area, or the person moves abnormally, it is determined to be abnormal behavior.

[0031] S23: Identify abnormal working environment status based on the collected video data of the power monitoring area.

[0032] In this embodiment, the specific steps of identifying the abnormal working environment state of the collected video data of the power monitoring area are as follows: S231. Detect abnormal sounds in the environment through sound recognition technology; wherein, abnormal sounds include prolonged abnormal sounds from power facilities and loud sounds in the environment.

[0033] S3. Match the extracted abnormal events with the complete video data of the power monitoring area saved in the blockchain.

[0034] In this embodiment, the extracted abnormal events are matched with the complete video data of the power monitoring area saved in the blockchain. The specific steps are as follows: S31, for the video of the identified abnormal environment, matching the matched complete video with the complete video around the work scene in the same time period, and according to the start and end timestamps of the video segment A, extracting the sub-segment of the same time period as the video segment A from other complete videos around the work scene to generate the video segment B, and combining the video segment A and the video segment B into a video segment to form a multi-view video set; S32. For videos showing abnormal identification of personnel and when personnel are not working in the standard area, an index is created for each identified object based on the identified facial information, and all related videos showing the target person in video segment A appearing within a set time window are searched in the complete video data based on the created object index. The video capture time is set, video segments are generated, and videos showing abnormal identification of personnel are compiled into a video set to form a video set showing abnormal identification of personnel.

[0035] S4. Retrieve the video data corresponding to the abnormal event on the blockchain based on the matching results.

[0036] Embodiment 2 like Figure 2As shown, this embodiment provides a traceability system for on-site power supervision service behavior based on blockchain, including a video acquisition module, a video transmission module, a video data classification module, a compression module, a receiving module, a blockchain server a, a blockchain server b and a traceability platform, the traceability platform includes an abnormal event extraction module, an event matching module, a video acquisition module, a video transmission module, a video data classification module, a compression module, a receiving module, and a blockchain server a are in communication connection, and the abnormal event extraction module and event matching module of the traceability platform are in communication connection with the receiving module, the blockchain server b, and the blockchain server a; Video acquisition module: confirm the camera video monitoring range, and collect video data of staff service behaviors and on-site work scenes in real time; In this embodiment, the video acquisition module: confirms the camera video monitoring range, and collects the video data of the staff's service behavior and the on-site work scene in real time as follows: Confirm the monitoring range of the camera to ensure full coverage and avoid blind spots or dead angles; The camera begins to collect video data of the service behaviors of on-site staff and on-site work scenes.

[0037] Video transmission module: transmits the collected video data through network media and stores it through the blockchain network; It should be noted that the specific steps of camera installation are: pre-install high-definition cameras in the power facility area to ensure full coverage and enable the cameras to capture video data of staff service behaviors and on-site work scenes with high resolution: Carry out layout planning in the power facility area, determine the installation location of the camera, and consider the coverage requirements of key areas and key locations to ensure a comprehensive monitoring range; According to the layout plan, install high-definition cameras at predetermined locations to ensure that the cameras can effectively cover the required monitoring area, including the working area and the surrounding environment; After installation, the camera is debugged and tested to ensure that it can work properly and collect video footage of the service behavior of on-site staff and on-site work scenes; Confirm the monitoring range of the installed cameras, check whether there are blind spots or dead angles, and make adjustments or additional installations as needed to ensure full coverage of the monitoring range; The camera begins to collect video data of the service behaviors of on-site staff and on-site work scenes.

[0038] Through layout planning and camera installation, we can ensure adequate coverage and comprehensive monitoring, including key areas and key locations, thereby improving the effectiveness and reliability of the monitoring system. At the same time, the staff's behavior and on-site conditions are clearly recorded, which helps to optimize supervision management and work flow.

[0039] In this embodiment, the video transmission module transmits the collected video data through the network medium and stores it through the blockchain network as follows: Detect and test network configuration to ensure the normal operation of network equipment, and perform bandwidth management to ensure sufficient bandwidth for video transmission; Select the transmission protocol according to actual needs and network environment; Video data is divided into multiple small data packets for transmission through streaming media transmission technology.

[0040] Video data classification module: used to classify video data according to time; Compression module: compresses the classified video data to reduce the data volume to improve storage and transmission efficiency, and finally transmits the compressed video data.

[0041] It should be noted that by configuring and testing the network, the network can be optimized according to the needs of video transmission. At the same time, potential problems in the network can be discovered and solved, such as network congestion, insufficient bandwidth, equipment failure, etc., thereby improving the stability and reliability of the network, reducing failures and interruptions during video transmission, reducing the time and cost of subsequent debugging and repair, and improving the efficiency and reliability of video transmission.

[0042] In this embodiment, the compression module compresses the classified video data to reduce the data volume to improve storage and transmission efficiency, and finally transmits the compressed video data in the following process: Compressing video data classified according to time; Set different compression encoding parameters for video data at different times; The classified video data is compressed and encoded to convert the video data into a compressed format to reduce the amount of data.

[0043] It should be noted that classifying by time can improve the readability of video data. Classifying and storing videos by time period can make the data orderly arranged on the time axis, and the data structure is clearer, which can also make subsequent management and analysis work more efficient. Compression processing can significantly reduce the size of video data, thereby saving storage space and reducing transmission costs.

[0044] Abnormal event extraction module: extracts the video of the staff performing on-site power service behaviors, including: personnel identity recognition, abnormal behavior recognition, abnormal working environment status recognition, and determines the advance and delay of video time periods based on the different abnormal events recognized.

[0045] In this embodiment, the abnormal event extraction module extracts the video of the staff performing on-site power service behaviors, including: personnel identity recognition, abnormal behavior recognition, abnormal working environment status recognition, and judging the advance and delay of the video time period according to the different abnormal events recognized. The processing process is as follows: Personnel identification: Combined with face recognition technology, identity verification of on-site staff and check their work qualifications: Use facial recognition technology to detect and identify the collected facial data to determine the identity of the person, and compare the identified facial data with a known staff database to verify the identity of the on-site staff; When identity recognition fails, the video is marked with an abnormal identity label for the person, and the identity of the person is predicted by integrating multiple features including body shape features, external work clothes features, voiceprint features, and hair features: The body shape features, external work clothes features, voiceprint features and hair features of the people in the video are extracted, and the four extracted features are weighted and integrated. The specific formula is: ; in, is the result of weighted fusion of all modal features. They are the weight of body shape feature, the weight of work clothes feature, the weight of voiceprint feature, and the weight of hair feature; The fused feature vector Input into the classifier to obtain the original output value of the classifier, use the Softmax function to normalize the original output value to obtain the predicted probability distribution, and from the probability distribution, select the three results with the highest probability and their corresponding category indexes. Finally, obtain the corresponding category and probability value and output them for the supervisor to judge the identity of the person.

[0046] It should be noted that in the case of insufficient light or when the face is blocked, the identification of a person may fail. Therefore, by integrating multiple features to predict the identity of a person, the limitation of single feature identification can be overcome, and the accuracy and reliability of identity prediction can be significantly improved, so that the identity of a person can be more effectively locked in complex environments or special circumstances, making it easier for supervisors to quickly determine the identity of a person. When the identity recognition is successful, the staff member is further checked for work qualifications; Match the type of supervision service with the on-the-job staff with corresponding qualifications. If the match is successful, it means that the qualification verification is successful, and then the service behavior of the staff is continuously monitored; When qualification verification fails, the collected facial data will be matched with all on-the-job personnel to identify the specific identity of the staff member, and the identified specific qualification information will be marked on the staff member. At the same time, an abnormal label of the work qualification of the person will be marked in the video.

[0047] In this embodiment, the abnormal behavior identification process is as follows: Monitoring of abnormal behaviors of staff members, including situations where staff members wear safety equipment, work in abnormal areas, move abnormally, or stay abnormally: Using object detection algorithms, train models to identify safety equipment such as safety gloves, safety glasses, and masks; Extract feature representation of the image through convolutional neural network, apply object detection algorithm on the feature map, identify the security device area in the image through bounding box regression and classifier, and output the identified security device and location information; When the target detects that a worker is not wearing the necessary safety equipment to perform on-site operations, it is considered abnormal behavior; The specific formula for determining the working position of the staff is: ; Among them, T is the number of intersection points, P is the point where the staff member is currently located, is the boundary segment of the specified area, intersect Indicates judgment point Is it consistent with the boundary segment? The intersection function, For all boundary segments ( ) to obtain the total number of intersections between all boundary segments and point (P); An even number of intersections indicates that the worker is outside the specified area. When the worker is outside the specified area, it is determined to be working in an abnormal area; An odd number of intersections indicates that the staff is working in the specified area; The next step of testing will be carried out after the staff is detected to be in the specified area; Use anomaly detection algorithms to analyze video data and detect abnormal situations in a timely manner, including abnormal movement and abnormal stay; The specific formula is: ; in, and represents the position of the object in the current frame (t), and Indicates the position of the object in the previous frame (t-1). By calculating the Euclidean distance between the current frame and the previous frame, the position change S of the object can be obtained; The threshold range of normal operation is set to [(M-1), (M+1)]. When S < (M-1), the worker is judged to be abnormally staying; when S > (M+1), the worker is judged to be abnormally moving; Identification of abnormal working environment status: When workers are performing power service activities, they use image analysis and deep learning technology to identify abnormal working environments, including the presence of flames and smoke in the environment; When workers are performing electricity services, they use sound recognition technology to detect abnormal sounds in the environment, including prolonged abnormal noises from power facilities and loud sounds in the environment.

[0048] In this embodiment, the process of determining the advance and delay of the video time period according to the different abnormal events identified is as follows: Personnel identification anomalies include personnel identity anomalies and personnel work qualification anomalies. When personnel identification anomalies occur, the data video of the personnel entering the work area and the video data of the personnel leaving the work area are extracted; When abnormal behavior recognition occurs, if the staff is in a non-specified area, the video data of the staff's entry and exit time in the non-specified area will be extracted; if the staff is working in a specified area, the video data of the staff's abnormal movement and abnormal stay will be extracted; When an abnormality occurs in abnormal environment recognition, video data of the abnormal environment is extracted, including data of environmental abnormalities and abnormal sounds.

[0049] The time for video extraction is set according to these three situations. When personnel identification is abnormal or personnel are not working in the standard area, the recorded personnel entry time and exit time are M and D respectively, and the video data of (M-1, M+1) and (D-1, D+1) are extracted.

[0050] Abnormal movement and abnormal stay of personnel: The time starting point of abnormal movement and stay of personnel is W, and the video data of (W-5, W+5) is extracted.

[0051] Abnormal environment recognition: The time when the abnormality occurs in the environment is K, and the video data of (K-10, K+10) is extracted.

[0052] It should be noted that since the video data will only be judged as abnormal data after the abnormal event is detected, it is impossible to clearly understand the beginning and end of the abnormal event. Therefore, the time of abnormal video data is advanced and delayed according to different abnormal events, and extracted so that supervisors can have a clearer understanding of the abnormal event when querying and retrieving it.

[0053] Event matching module: Match the extracted video clips of abnormal events with the complete video. When the abnormal video clips successfully match the video data of the complete event, the complete video is extracted and further automatically advanced and extended.

[0054] In this embodiment, the event matching module: matches the extracted video clip of the abnormal event with the complete video. When the abnormal video clip matches the video data of the complete event successfully, the complete video is extracted, and further automatic advance and extension processing is performed as follows: The extracted video data of abnormal situations in the three working scenes are matched with the complete video through the database; Use the timestamp matching method to automatically compare the timestamp of the extracted abnormal situation video clip with the timestamp of the complete video in the database to find the exact time point when the abnormal event occurred; When the video is successfully matched with the complete video, the video time is further automatically advanced and extended to generate video segment A; For the video of the identified abnormal environment, the matched complete video is matched with the complete video around the work scene in the same time period. According to the start and end timestamps of video segment A, the sub-segments of the same time period as video segment A are cut out from other complete videos around the work scene to generate video segment B. Video segment A and video segment B are combined into video segments to form a multi-view video set for the supervisor to view; For videos with abnormal identification of personnel and when personnel are not working in the standard area, an index is created for each identified object based on the identified face information, and all relevant videos appearing within a 12h window with the target personnel in video clip A are searched in the complete video data based on the established object index, the video capture time is set, video clips are generated, and videos with abnormal personnel are compiled into a video set to form a video set with abnormal personnel identities for the supervisor to review; For persons whose specific identities are not recognized, based on the three results of identity prediction, all relevant videos in which the predicted person appears within a 12h window are searched in the complete video data, the video capture time is set, video clips are generated, and videos in which the predicted person appears abnormal are made into video sets, thus forming three video sets with abnormal identities of the predicted person for supervisory personnel to review.

[0055] It should be noted that by extracting the complete video around the work scene, the same event can be observed from multiple camera perspectives, providing a more comprehensive event restoration. Multiple camera perspectives help analyze the environmental conditions and personnel flow of the event, and enhance the comprehensive understanding of the event.

[0056] The overall time of the abnormal event period is extended, and the extended video is automatically displayed first. At the same time, since the complete video has been matched, the advance and delay time can also be manually adjusted.

[0057] It should be noted that by advancing and extending the video data, a more complete and comprehensive development process of abnormal events can be obtained. At the same time, this helps to gain a deeper understanding of the beginning and end of the event, including the cause, process and result of the event, which is conducive to making more comprehensive analysis and judgment.

[0058] Receiving module: It is a computer terminal used to view, retrieve, manage and copy the video data of on-site power supervision service behaviors. It is connected to blockchain server a and blockchain server b. The complete video data is stored in blockchain server a, and the video data of the traceability platform is stored in blockchain server b. After confirmation, the video data will be stored and the transaction will be permanently recorded on blockchain server a and blockchain server b, forming an irreversible historical record.

[0059] In this embodiment, the receiving module is a computer terminal, which is used to view, retrieve, manage and copy the video data of the on-site power supervision service behavior, and is connected to the blockchain server a, the blockchain server b and the database. The complete video data is stored in the blockchain server a, the video data of the traceability platform is stored in the blockchain server b, and the extracted abnormal events are stored in the database. After the video data is confirmed, the transaction will be permanently recorded on the blockchain server a and the blockchain server b, and the process of forming an irreversible historical record is as follows: The complete video data is transmitted to a computer terminal, and the video data is submitted to the blockchain server a through the computer terminal. The complete video data is uploaded to the off-chain storage system using the off-chain storage method, and the hash value of the video data is further calculated. The hash value and related metadata are then submitted to the preset smart contract a. Smart contract a receives the hash value and metadata of the uploaded data and verifies the data. After the verification is passed, the hash value and related metadata of the video data are recorded on the blockchain server a; The video data of the traceability platform is submitted to the blockchain server b, which is preset with a smart contract b. Smart contract b verifies the intercepted data hash value and metadata on the blockchain server b, and matches and verifies the traceability data with the record of the complete video data. After the verification is passed, the traceability data and its hash value and metadata are recorded on the blockchain server b and stored in the chain storage method; After the data is recorded on the blockchain server, it forms an irreversible historical record; Supervisors retrieve, view and copy stored video data through computer terminals.

[0060] It should be noted that storing video data through blockchain can form an irreversible historical record, ensure the security and integrity of the data, and effectively prevent malicious tampering and destruction of the data. At the same time, setting up blockchain server a and blockchain server b makes video data tracing more convenient.

[0061] In summary, by setting up blockchain server a and blockchain server b, blockchain server a is used to store complete operation videos, and blockchain server b is used to store videos of the traceability platform, so that the two types of videos can be classified and stored. At the same time, due to limited storage space, the video data that can be processed by the traceability platform is smaller than the complete video data, so the stored data range is wider and the traceability time is longer; at the same time, by storing data on blockchain server a and blockchain server b, since the data on blockchain server a and blockchain server b are saved in an encrypted and distributed storage manner, once the data is stored on server a and blockchain server b and confirmed, it cannot be tampered with or deleted, thereby ensuring The integrity and security of data can be ensured, data can be prevented from being maliciously tampered with or damaged, and the tracing time can be increased; by setting up an abnormal event extraction module, abnormal event fragments can be extracted from the complete video data, and the traditional abnormal event video extraction is usually extracted after the abnormal event occurs, which may cause the intercepted video fragments to be missing. Therefore, an abnormal event extraction module is set to intercept the video in advance and delay, and at the same time, abnormal events are classified into three situations: personnel identity recognition, abnormal behavior recognition, and abnormal working environment status recognition. According to these three different situations, the time is advanced and delayed respectively, so that the intercepted video duration in the three situations is different, which is convenient for viewing. When the specific identity information of a person cannot be recognized through facial data, the body shape features, external work clothes features, voiceprint features and hair features of the person in the video will be extracted, and the extracted features will be integrated to predict the identity of the person; by setting up an event matching module, the next step of video matching will be automatically carried out according to the known environmental abnormal event video. After the video is successfully matched with the complete video data, the video time will be further extended to generate video A. The complete video data will be timestamped with the complete video around the work scene extracted from the database to generate video clip B. The video collection generated by combining video A and video clip B not only enables the supervisor to understand the whole event more clearly, but also enables the supervisor to use multiple The same event can be observed from different camera perspectives, thus providing a more comprehensive and three-dimensional restoration of the event. In addition, after the matching is completed, the length of the captured video can be manually adjusted to make tracing simpler and convenient for supervision. According to the known videos of abnormal personnel identity events, all relevant videos of the target person appearing within a 12h window will be captured and compiled according to the recognized facial information to form a collection of videos of abnormal personnel identity for review by supervisory personnel. For persons whose specific identities are not recognized, according to the three results of identity prediction, all relevant videos of the predicted person appearing within a 12h window will be searched in the complete video data to form three video collections of predicted person identity abnormalities for review by supervisory personnel to further fully understand the whole story.

[0062] In further embodiments, there is also provided: An electronic device includes a memory and a processor, and computer instructions stored in the memory and executed on the processor, wherein when the computer instructions are executed by the processor, the method in Embodiment 1 is performed. For the sake of brevity, it will not be described in detail here.

[0063] It should be understood that in this embodiment, the processor may be a central processing unit CPU, and the processor may also be other general-purpose processors, digital signal processors DSP, application-specific integrated circuits ASIC, off-the-shelf programmable gate arrays FPGA or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0064] The memory may include a read-only memory and a random access memory, and provide instructions and data to the processor. A portion of the memory may also include a non-volatile random access memory. For example, the memory may also store information about the device type.

[0065] A computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, the method in embodiment 1 is completed.

[0066] The method in the first embodiment can be directly embodied as a hardware processor, or a combination of hardware and software modules in the processor. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware. To avoid repetition, it will not be described in detail here.

[0067] A computer program product includes a computer program, and when the computer program is executed by a processor, the method in the first embodiment is implemented.

[0068] The present invention also provides at least one computer program product tangibly stored on a non-transitory computer-readable storage medium. The computer program product includes computer executable instructions, such as instructions included in a program module, which is executed in a device on a real or virtual processor of the target to perform the above process / method. Typically, a program module includes routines, programs, libraries, objects, classes, components, data structures, etc. that perform specific tasks or implement specific abstract data types. In various embodiments, the functions of the program modules can be combined or divided between program modules as needed. The machine executable instructions for the program modules can be executed in local or distributed devices. In distributed devices, the program modules can be located in local and remote storage media.

[0069] The computer program code for implementing the method of the present invention can be written in one or more programming languages. These computer program codes can be provided to the processor of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the computer or other programmable data processing device, causes the function / operation specified in the flow chart and / or block diagram to be implemented. The program code can be executed completely on a computer, partially on a computer, as an independent software package, partially on a computer and partially on a remote computer or completely on a remote computer or server.

[0070] In the context of the present invention, computer program codes or related data may be carried by any appropriate carrier to enable a device, apparatus or processor to perform the various processes and operations described above. Examples of carriers include signals, computer readable media, etc. Examples of signals may include electrical, optical, radio, acoustic or other forms of propagation signals, such as carrier waves, infrared signals, etc.

[0071] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with this embodiment can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0072] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. A blockchain-based traceability method for on-site power supervision service behavior, characterized in that: include: S1. Collect video data of the power monitoring area and save it to the blockchain; S2. Extracting abnormal events from the collected video data of the power monitoring area; wherein the abnormal events include abnormal personnel identity, abnormal behavior and abnormal working environment status; S3, matching the extracted abnormal events with the complete video data of the power monitoring area saved to the blockchain; S4. Retrieve the video data corresponding to the abnormal event on the blockchain according to the matching result.

2. The method for tracing on-site power supervision service behavior based on blockchain as claimed in claim 1, characterized in that: The abnormal events are extracted from the collected video data of the power monitoring area; wherein the abnormal events include abnormal personnel identity, abnormal behavior and abnormal working environment state, and the specific steps are: S21, performing personnel identification on the collected video data of the power monitoring area; S22, identifying abnormal behavior on the collected video data of the power monitoring area; S23: Identify abnormal working environment status based on the collected video data of the power monitoring area.

3. The method for tracing on-site power supervision service behavior based on blockchain as claimed in claim 2 is characterized in that: The specific steps of performing personnel identification on the collected video data of the power monitoring area are as follows: S211, extracting body shape features, external work clothes features, voiceprint features and hair features of the people in the video; S212, performing weighted fusion on the extracted features to obtain fusion features of the people in the video; S213: Input the fusion features of the persons in the video into the trained classifier to obtain the identity recognition results of the persons in the video.

4. The method for tracing on-site power supervision service behavior based on blockchain as claimed in claim 2, characterized in that: The specific steps of performing abnormal behavior recognition on the collected video data of the power monitoring area are as follows: S221, extracting image feature representation in the video through a convolutional neural network; S222, using a target detection algorithm to identify whether a person in the video image is wearing safety equipment through bounding box regression and a classifier; S223, determining whether the person is in the specified working area according to the relationship between the position of the person in the video and the specified working area; The specific formula for determining whether a person is in the specified working area is as follows: ， Among them, T is the number of intersection points, P is the point of the current position of the staff, is the boundary line segment of the specified area, intersect represents the function of judging whether the point intersects with the boundary line segment, and represents the sum of all boundary line segments to obtain the sum of the number of intersections between all boundary line segments and point P; when the number of intersection points is an even number, it means that the staff is outside the specified area. When the staff is outside the specified area, it is determined to be working in an abnormal area; when the number of intersection points is an odd number, it means that the staff is working in the specified area; S224, using an anomaly detection algorithm to detect the movement of people in the video, and determine whether there is abnormal movement of people; S225. If it is detected that at least one of the following situations occurs: the person is not wearing safety equipment, the person is not in the specified working area, or the person moves abnormally, it is determined to be abnormal behavior.

5. The method for tracing on-site power supervision service behavior based on blockchain as claimed in claim 4 is characterized in that: The abnormal detection algorithm is used to detect the movement of people in the video to determine whether there is abnormal movement of people, specifically: S2241, obtaining a position change of the staff member by calculating the Euclidean distance according to the position of the staff member in the current frame and the position of the staff member in the previous frame; S2242. Determine whether the worker has moved abnormally based on the worker's position change and the set threshold range of normal operation.

6. The method for tracing on-site power supervision service behavior based on blockchain as claimed in claim 2, characterized in that: The specific steps of identifying the abnormal working environment state of the collected video data of the power monitoring area are as follows: Abnormal sounds in the environment are detected through sound recognition technology; wherein, the abnormal sounds include abnormal sounds of power facilities that last for more than a set time and sounds in the environment that exceed a set volume.

7. The method for tracing on-site power supervision service behavior based on blockchain as claimed in claim 1, characterized in that: The extracted abnormal events are matched with the complete video data of the power monitoring area saved in the blockchain, and the specific steps are as follows: S31, for the video of the identified abnormal environment, matching the matched complete video with the complete video around the work scene in the same time period, and according to the start and end timestamps of the video segment A, extracting the sub-segment of the same time period as the video segment A from other complete videos around the work scene to generate the video segment B, and combining the video segment A and the video segment B into a video segment to form a multi-view video set; S32. For videos showing abnormal identification of personnel and when personnel are not working in the standard area, an index is created for each identified object based on the identified facial information, and all related videos showing the target person in video segment A appearing within a set time window are searched in the complete video data based on the created object index. The video capture time is set, video segments are generated, and videos showing abnormal identification of personnel are compiled into a video set to form a video set showing abnormal identification of personnel.

8. The blockchain-based on-site power supervision service behavior traceability system is characterized by: include: Video acquisition module, used to collect video data of the power monitoring area and save it to the blockchain; An abnormal event extraction module is used to extract abnormal events from the collected video data of the power monitoring area; wherein the abnormal events include abnormal personnel identity, abnormal behavior and abnormal working environment status; An event matching module, used to match the extracted abnormal events with the complete video data of the power monitoring area saved to the blockchain; The tracing platform is used to retrieve the video data corresponding to the abnormal event on the blockchain based on the matching results.

9. The blockchain-based on-site power supervision service behavior tracing system as claimed in claim 8, characterized in that: The abnormal event extraction module includes: A personnel identification module is used to identify personnel from the collected video data of the power monitoring area; An abnormal behavior recognition module is used to recognize abnormal behaviors of the collected video data of the power monitoring area; The abnormal working environment status module is used to identify the abnormal working environment status of the collected video data of the power monitoring area.

10. The blockchain-based on-site power supervision service behavior tracing system as claimed in claim 8, characterized in that: The event matching module includes: Construct a multi-view video set module, which is used to match the matched complete video with the complete video around the work scene in the same time period for the video of the identified abnormal environment, and extract the sub-segment of the same time period as the video segment A from other complete videos around the work scene according to the start and end timestamps of the video segment A to generate the video segment B, and combine the video segment A and the video segment B into a video segment to form a multi-view video set; Construct a video collection module for abnormal personnel identity, which is used for videos of abnormal personnel identity recognition and personnel not working in the standardized area. According to the recognized facial information, an index is established for each identified object, and all related videos appearing in the target person in video segment A within the set time window are searched in the complete video data according to the established object index. The video capture time is set, and video segments are generated. The videos with abnormal personnel are made into a video collection to form a video collection of abnormal personnel identity.

Images