阅读说明：本技术 一种基于高精度定位和图像识别的视频告警联动实现方法 (Video alarm linkage implementation method based on high-precision positioning and image recognition ) 是由 王学广 刘金锁 彭启伟 李国志 樊卫东 陈坤鹏 郝小龙 郭佳 徐濛 张军 彭家陵 于 2021-07-27 设计创作，主要内容包括：本发明公开了一种基于高精度定位和图像识别的视频告警联动实现方法,边缘计算装置将定位安全帽的定位信息与RFID标签信息进行关联绑定；边缘计算装置根据指定定位安全帽的定位信息与对应上传的RFID标签信息进行关联绑定,获取电子围栏节点的定位信息及RFID标签信息；将电子围栏节点位置信息自动围合形成闭合空间；边缘计算装置将布控球默认预置位作为联动基准位置；对作业人员违章行为进行判别,并进行告警提醒。本发明提供的一种基于高精度定位和图像识别的视频告警联动实现方法,满足现场告警、实时跟踪与取证要求。(The invention discloses a video alarm linkage realization method based on high-precision positioning and image recognition.A border computing device associates and binds positioning information of a positioning safety helmet and RFID label information; the edge computing device performs association binding according to the positioning information of the appointed positioning safety helmet and the correspondingly uploaded RFID label information to acquire the positioning information and the RFID label information of the electronic fence node; automatically enclosing the position information of the electronic fence node to form a closed space; the edge computing device takes the default preset position of the control ball as a linkage reference position; and (4) judging the violation behaviors of the operating personnel and giving an alarm for reminding. The video alarm linkage implementation method based on high-precision positioning and image recognition provided by the invention meets the requirements of on-site alarm, real-time tracking and evidence obtaining.)

1. A video alarm linkage realization method based on high-precision positioning and image recognition is characterized in that: the method comprises the following steps:

the code scanning gun scans the RFID label of the positioning safety helmet and sends the RFID label information to the edge computing device, and the edge computing device associates and binds the positioning information of the positioning safety helmet and the RFID label information;

a person wearing the positioning safety helmet specifies positioning information of the positioning safety helmet through an APP in a mobile phone, carries a code scanning gun to scan the RFID label on the electronic fence node, and sends the RFID label information to an edge computing device, and the edge computing device performs association binding with the correspondingly uploaded RFID label information according to the specified positioning safety helmet positioning information to obtain the electronic fence node positioning information and the RFID label information;

the edge computing device calls a map layer interface according to the APP, the position information of the electronic fence nodes is automatically enclosed to form a closed space, the starting point and the end point of the scanned electronic fence are used as an entrance, and the closed space of the electronic fence is divided into areas according to needs;

the edge computing device acquires the position information of the distribution control ball, controls the distribution control ball to rotate through a control protocol, realizes identification of the label through an object identification algorithm, places the label on a picture mark position through image analysis and processing, sets a default preset position of the distribution control ball through the control protocol, and takes the default preset position of the distribution control ball as a linkage reference position;

the edge computing device judges the violation behaviors of the operating personnel by acquiring the positioning data and the electronic fence information of the operating personnel wearing the positioning safety helmet and carries out alarm reminding.

2. The video alarm linkage implementation method based on high-precision positioning and image recognition as claimed in claim 1, wherein: the areas are set as a working area and a danger area.

3. The video alarm linkage implementation method based on high-precision positioning and image recognition as claimed in claim 1, wherein: the control protocol adopts an ONVIF protocol.

4. The video alarm linkage implementation method based on high-precision positioning and image recognition as claimed in claim 1, wherein: and the picture zone bit is set as a picture horizontal central point.

5. The video alarm linkage implementation method based on high-precision positioning and image recognition as claimed in claim 1, wherein: the violation behaviors include: 1) the method comprises the following steps that (1) an operator enters a dangerous area and is judged and realized in an electronic fence area through personnel positioning data, 2) the operator does not enter the electronic fence from an entrance and an exit, and is judged and realized through comparison of personnel positioning data historical tracks and electronic fence closed space relations, 3) the operator leaves the electronic fence working area for more than n minutes, a timing function is activated outside the electronic fence working area through the personnel positioning data, and is judged and realized for more than n minutes, and 4) the operator climbs a tower and ascends the tower, and is judged and realized through personnel elevation data and more than an elevation threshold value; the alert reminder includes: when the violation behaviors are met, the edge calculating device controls the distribution control ball to rotate to the linkage reference position through the control protocol, the rotation angle theta of the distribution control ball is calculated according to the collected violation worker positioning information and the linkage reference position, the steering angle is calculated through theta (1/180), and the picture of the violation worker is rotatably positioned by controlling the relative movement angle of the cloud deck.

6. The video alarm linkage implementation method based on high-precision positioning and image recognition as claimed in claim 1, wherein: further comprising: the edge computing device carries out image analysis and identification on the video picture, adjusts the focal length to enable an operator to be in the middle of the picture, simultaneously carries out violation video screenshot and video evidence collection, calls a face recognition algorithm to determine the identity of the violation person, sends detailed alarm information to a supervisor, and the supervisor can carry out remote video picture monitoring.

7. The video alarm linkage implementation method based on high-precision positioning and image recognition as claimed in claim 1, wherein: further comprising: after the alarm occurs, if the violation operation personnel walks, the edge computing device starts an automatic tracking algorithm of the object, controls the ball distributing and controlling holder to automatically turn to track the violation personnel, and is convenient for the supervision personnel to dynamically monitor the action behavior of the violation personnel in real time.

8. The video alarm linkage implementation method based on high-precision positioning and image recognition as claimed in claim 1, wherein: as an optimal scheme, a Beidou positioning module is adopted as a positioning module of the positioning safety helmet and is used for acquiring position data and elevation data.

9. The video alarm linkage implementation method based on high-precision positioning and image recognition as claimed in claim 1, wherein: the system further comprises a cloud management platform, wherein the cloud management platform maintains the edge computing device ledgers of a plurality of operation sites and maintains the bound RFID codes for a plurality of positioning safety caps and electronic fence nodes.

10. The video alarm linkage implementation method based on high-precision positioning and image recognition as claimed in claim 1, wherein: further comprising: when the position of the distribution control ball moves, the edge computing device acquires the positioning information of the distribution control ball again, the distribution control ball identifies the label again, and the default preset position of the distribution control ball is reset.

Technical Field

The invention relates to a video alarm linkage realization method based on high-precision positioning and image recognition, and belongs to the technical field of security intelligent alarm.

Background

In professional operation construction sites of power transformation, power transmission, power distribution and the like, safety supervision needs to be carried out on the actions of operators, such as breaking into operation dangerous areas, climbing towers and ascending heights, and in the traditional method, supervisors need to be invested to carry out site real-time supervision.

In the prior art, a pulse electronic fence is adopted to form an operation area, hardware equipment needs to be arranged on site, an alarm is generated by using an electronic signal, the directional intrusion detection is realized, accurate positioning information cannot be provided, and the problem of inconvenience in power supply exists. Because the field ball distribution and control monitoring visual angle has direction limitation and a monitoring blind area, if high-precision positioning data does not exist, the problems of steering delay and position accuracy when the ball distribution and control monitoring visual angle is controlled to be turned to the position of a violation worker are solved when an alarm is generated, and effective evidence monitoring picture loss is caused.

Disclosure of Invention

The purpose is as follows: in order to overcome the defects in the prior art, the invention provides a video alarm linkage implementation method based on Beidou positioning and image recognition, which is characterized in that a digitalized electronic fence is implemented by utilizing a Beidou high-precision positioning technology, and the violation behaviors such as no safety helmet, no safety belt and the like and the recognition of reference points are recognized by utilizing image analysis, so that the control ball can be automatically and timely controlled to be transferred to the position of a violation worker according to high-precision positioning data when violation occurs, and the tracking and evidence obtaining requirements are met.

The technical scheme is as follows: in order to solve the technical problems, the technical scheme adopted by the invention is as follows:

a video alarm linkage realization method based on high-precision positioning and image recognition comprises the following steps:

the code scanning gun scans the code of the RFID label of the positioning safety cap and sends the information of the RFID label to the edge computing device, and the edge computing device associates and binds the positioning information of the positioning safety cap and the information of the RFID label.

The personnel who wear the location safety helmet appoint the locating information of this location safety helmet through APP in the cell-phone to carry and sweep the sign indicating number rifle and sweep the sign indicating number to the RFID label on the fence node, and give edge computing device with RFID label information transmission, edge computing device carries out the relevance according to the locating information of appointed location safety helmet and the RFID label information that corresponds and upload and binds, acquires the locating information and the RFID label information of fence node.

The edge calculating device calls a map layer interface according to the APP, the electronic fence node position information is automatically enclosed to form a closed space, the scanned electronic fence starting point and the scanned electronic fence ending point serve as an entrance and an exit, the closed space of the electronic fence is divided into a working area and a danger area as required, and the digital electronic fence is drawn.

The edge computing device obtains position information of the cloth control ball, controls the cloth control ball to rotate and lift through an ONVIF protocol, realizes identification and judgment of the label through calling an object identification algorithm, carries out image analysis processing and places the label on a picture mark position such as a picture horizontal central point, and sets a default preset position of the cloth control ball through the ONVIF protocol to serve as a linkage reference position.

The edge computing device judges the behavior of the operator by acquiring the positioning data and the electronic fence information of the operator wearing the positioning safety helmet and carries out violation behavior warning reminding; the violation behaviors include: 1) the method comprises the following steps that (1) an operator enters a dangerous area and is judged and realized in an electronic fence area through personnel positioning data, 2) the operator does not enter the electronic fence from an entrance and an exit, and is judged and realized through comparison of personnel positioning data historical tracks and electronic fence closed space relations, 3) the operator leaves the electronic fence working area for more than n minutes, a timing function is activated outside the electronic fence working area through the personnel positioning data, and is judged and realized for more than n minutes, and 4) the operator climbs a tower and ascends the tower, and is judged and realized through personnel elevation data and more than an elevation threshold value; when the violation behaviors are judged, the edge computing device controls the distribution control ball to rotate to the linkage reference position through the ONVIF protocol, the rotating angle theta of the distribution control ball is computed according to the collected violation worker positioning data and the linkage reference position, the steering angle is computed through theta (1/180), the relative movement angle of the cradle head is controlled through calling the ONVIF protocol, the relative lifting height of the cradle head is controlled through calling the ONVIF protocol according to the collected violation worker elevation data, and the distribution control ball is rapidly rotated and positioned to the violation worker picture.

The edge computing device carries out image analysis and identification on the video picture, adjusts the focal length to enable an operator to be in the middle of the picture, simultaneously carries out violation video screenshot and video evidence collection, calls a face recognition algorithm to determine the identity of the violation person, sends detailed alarm information to a supervisor, and the supervisor can carry out remote video picture monitoring.

After the alarm occurs, the object automatic tracking algorithm of the edge computing device is started when the operator walks, the ball distributing and controlling holder is controlled to automatically turn to track the violation personnel, and the monitoring personnel can dynamically monitor the action behavior of the violation personnel in real time.

Preferably, the positioning safety helmet is provided with a positioning module.

As an optimal scheme, the positioning module adopts a Beidou high-precision positioning module which is used for acquiring position data and elevation data.

As a preferred scheme, the system further comprises a cloud management platform, wherein the cloud management platform maintains the edge computing device ledgers of a plurality of operation sites and maintains the bound RFID codes for a plurality of positioning safety caps and electronic fence nodes.

Has the advantages that: according to the video alarm linkage implementation method based on high-precision positioning and image recognition, a pulse fence is not needed, only an RFID label is configured at a node of an entity fence, a digital electronic fence is implemented by using an RFID technology and a Beidou high-precision positioning technology, local calculation is performed by using collected positioning data, elevation data and image analysis and recognition of an operator, and the position of a ball distribution and control steering violation operator can be timely controlled according to the high-precision positioning data when violation occurs, so that the requirements of on-site alarm, real-time tracking and evidence obtaining are met.

Drawings

FIG. 1 is a schematic diagram of a system architecture used in the method of the present invention.

FIG. 2 is a schematic flow chart of the method of the present invention.

Detailed Description

The present invention will be further described with reference to the following examples.

As shown in fig. 1, a video alarm linkage implementation system based on high-precision positioning and image recognition includes: the device comprises an edge computing device with a high computing power AI module, a scanning gun, an RFID tag, a control ball, a positioning safety helmet, a mobile phone containing an APP, a cloud management platform and a positioning host.

And the edge computing device performs data interaction with the cloud management platform, the positioning host and the mobile phone through a mobile network.

The edge computing device performs data interaction with the distribution control ball and the mobile phone through the wireless local area network.

And the control ball is used for acquiring a real-time video.

The edge computing device carries out data interaction with the code scanning gun through Bluetooth.

And the code scanning gun reads data of the RFID label.

The positioning host acquires the positioning information of the positioning safety helmet.

The cloud management platform is deployed at a cloud platform end and used for carrying out unified management configuration on a plurality of field devices and providing data interfaces for other service systems.

The ground enhanced base station is used for Beidou navigation satellite network signals.

The Beidou navigation satellite is used for providing high-precision positioning information and elevation information for equipment provided with the Beidou positioning module.

The sign is used for linking the reference point of the reference position.

The edge computing device is used for realizing linkage logic algorithm processing through an internal high-computation-power AI analysis module and an object recognition algorithm, a face recognition algorithm and an object dynamic tracking algorithm.

And the mobile phone application APP realizes acquisition, alarm data display, equipment and system binding and configuration.

As shown in figure 2, when the video alarm linkage implementation method based on high-precision positioning and image recognition is used, RFID tags are configured on a solid fence and a positioning safety helmet terminal of an operation site, before the site operation is started, a configurator wears a Beidou positioning safety helmet to perform RFID code scanning binding on solid fence nodes through a code scanning gun, positioning information of the positioning safety helmet is associated with the position points of the RFID tags of the solid fence, and the position coordinate information of each RFID tag solid fence is endowed through scanning and binding of a plurality of RFID tags of the solid fence, so that a digital electronic fence is formed.

The operating personnel wear a safety helmet carrying a Beidou high-precision positioning module to realize longitude and latitude position and height monitoring of the operating personnel, the Beidou high-precision positioning equipment is configured on the control distribution ball along with the machine body to realize position monitoring of the control distribution ball equipment, and automatic correction of a linkage reference position can be realized when the control distribution ball moves; arranging control balls outside the electronic fence of the operation area to perform video real-time operation monitoring, and hanging a label in the operation area as a linkage reference position.

And the linkage service, the intelligent analysis algorithm and the mobile phone APP software in the edge agent device realize configuration operation and data and alarm display. The edge computing device realizes linkage logic algorithm processing through an internal high-computation-power AI analysis module and an object recognition algorithm, a face recognition algorithm and an object dynamic tracking algorithm. The intelligent analysis algorithm comprises algorithms such as object recognition, face recognition, object dynamic tracking and the like; the linkage service realizes logic judgment according to collected positioning and image data, video linkage is realized by combining an image analysis algorithm, warning information is pushed to a local APP application and a cloud platform main station in time, the warning information is sent to a supervisor in time through MQTT message subscription, and screenshot and video evidence obtaining are carried out simultaneously.

Example 1:

a video alarm linkage realization method based on high-precision positioning and image recognition comprises the following steps:

1) the cloud management platform maintains the edge computing device ledgers of a plurality of operation sites, maintains bound RFID codes for a plurality of safety helmets and fence node terminals, reduces site configuration work, and RFID tags are attached to site fences and positioning safety helmets.

2) The method comprises the steps that an edge computing device scans codes of RFID (radio frequency identification) tags of a positioning safety helmet through a code scanning gun, determines that a designated positioning terminal provides positioning data to draw a fence, a background linkage service acquires longitude and latitude data of the positioning terminal from a positioning host and pushes the longitude and latitude data to a mobile phone APP (application), binding of the positioning data of the RFID tags is realized by sequentially dotting a plurality of RFID tags of the fence, the APP calls a map layer interface to automatically enclose a closed space according to dotting position coordinates, a scanning starting point and a scanning end point serve as an entrance and an exit, and fence marks can be configured into a working area and a dangerous area as required; and after the configuration is completed, transmitting the polygon drawn by the APP to a linkage service program of the edge computing device, and finishing the drawing of the digital electronic fence.

3) After a start-up process is started, a distributed control ball initialization instruction is issued by a linkage service in the edge computing device, longitude and altitude data of the distributed control ball are collected, the linkage service controls the distributed control ball to rotate 360 degrees through an ONVIF protocol, identification judgment of a label is realized by calling an object identification algorithm, the label is placed at a picture horizontal central point through image analysis processing, and a distributed control ball default preset position is set through the ONVIF protocol and serves as a linkage reference position.

4) When the behavior of an operator intruding or climbing a tower and ascending is caused, the linkage service judges whether the violation of regulations is caused by collecting high-precision positioning data of a safety helmet worn by the operator and combining an electronic fence, and an alarm is given when the following conditions are met: 1) personnel enter a dangerous area and are judged in a fence area through personnel positions, 2) the personnel do not enter the fence from an entrance and exit and are searched and compared with the current fence relationship through personnel positioning historical tracks for judgment, 3) the personnel leave the fence of a working area for n minutes and are judged by comparing the previous time with the current fence relationship and judging that the time exceeds n minutes; when the alarm condition is met, the linkage service firstly controls the distribution and control ball to be turned to a preset default preset position (reference position) through an ONVIF protocol, the rotation angle theta of the distribution and control ball is calculated according to collected longitude and latitude positioning data of an operator and longitude and latitude data of an entrance label, the steering angle is calculated through theta (1/180), the relative movement angle of a cradle head is controlled through calling the ONVIF cradle head control protocol, the cradle head can be quickly rotated and positioned to the picture of the violation operator, the video picture is analyzed and identified, the focal length is adjusted, the operator is positioned in the middle of the picture, meanwhile, violation video screenshot and video evidence collection are carried out, a face recognition algorithm is called to determine the identity of the violation operator, detailed alarm information is sent to a supervisor, and the supervisor can carry out remote video picture monitoring.

5) After the alarm occurs, the object automatic tracking algorithm of the edge computing device is started when the worker walks, the ball distributing and controlling holder is controlled to automatically turn to and track the violation personnel, and the monitoring personnel can dynamically monitor the action of the violation personnel in real time.

6) And 3, randomly placing the defense placement position of the distributed control ball by utilizing high-precision positioning equipment configured by the distributed control ball along with the machine body, and reinitializing in the step 3 when the linkage service acquires that the data of the positioning equipment randomly configured by the distributed control ball is changed.

Example 2:

the high-precision positioning information adopts a Beidou RTK base station and a positioning host machine mode to acquire centimeter-level high-precision longitude and latitude and height data, and meets the requirement of field operation safety and real-time performance.

A configuration personnel carrying the high-precision positioning module utilizes the code scanning gun to scan the RFID label and perform association binding on positioning information and the RFID, so that the common RFID label is changed into the positioning label, the digital electronic fence is realized at low cost, and the positioning terminal and the edge computing device are also provided with the RFID label to realize the binding of the terminal identity.

When an operator wearing the Beidou positioning safety helmet breaks into a dangerous area or climbs a tower to ascend, video linkage control is carried out through acquiring high-precision positioning data of the safety helmet to distribute and control balls to turn violation personnel, the violation personnel are identified through a face identification algorithm built in an edge computing device, warning information of the violation personnel is pushed to a supervisor in time, and the situation of violation on site can be known in time through the supervisor convenient to operate.

When an operator wearing the Beidou positioning safety helmet breaks into a dangerous area or climbs a tower to ascend, the arrangement and control ball can be timely and accurately controlled according to positioning data to turn to a monitoring picture of the violation operator, the violation operation and evidence obtaining requirements are met by capturing images and recording the images, meanwhile, an object tracking algorithm built in the edge computing device is started, the arrangement and control ball is controlled to turn in real time after the violation operator moves, and the violation operator is guaranteed to be in the monitoring picture of the arrangement and control ball.

When the position of the control ball moves, the high-precision positioning module configured with the control ball along with the machine body is utilized, and when the linkage service algorithm detects that the position of the control ball changes, linkage position correction is automatically carried out again without reconfiguration.

The invention is provided with corresponding positioning equipment, video monitoring equipment and edge computing equipment on the operation site, can realize the centimeter-level positioning precision of a positioning object through Beidou high-precision positioning, realize the real-time monitoring of an on-site operation picture through the image monitoring of a control ball, realize the on-site positioning information acquisition and image acquisition through an edge computing device, and realize the rapid linkage and tracking record of violation alarm through the image analysis of operation safety behaviors and the computation of positioning data.

The above description is only of the preferred embodiments of the present invention, and it should be noted that: it will be apparent to those skilled in the art that various modifications and adaptations can be made without departing from the principles of the invention and these are intended to be within the scope of the invention.