The noninductive panorama cognitive method of packaged type based on 3D laser
Technical field
The invention belongs to technical field of data processing, and in particular to panorama sense that a kind of packaged type based on 3D laser is noninductive
Perception method.
Background technique
With expanding rapidly for power grid scale, electric system scientific and technological level is continuously improved, and unmanned substation's quantity is increasingly
More, especially electric power vertically changes deepening continuously and optimizing for management system construction, administration needed for power generation/operation control personnel
Range it is increasing, this brings unprecedented challenge to power generation/operation control personnel.
In order to further increase substation's control means, using a variety of high accuracy positionings, perception Detection Techniques, development is based on
The application extension of the grid equipment centralized management platform of three-dimensional live technology makes power grid perception fusion intelligent ecological circle, in conjunction with
Existing OMS system, SCADA system building interactivity is good, i.e. visual sense is strong, immersion, true accurate power grid three-dimensional live intelligence
It can control platform.
Substation field charging equipment stands in great numbers, and lacks space exploration sense to make up traditional monitoring means to substation field
Know ability, using Airborne Lidar survey technology, researchs and develops device and technological means to operating area space exploration ability, knot
Close the three-dimensional live scene with space coordinate that restore to substation's perfect clone, realization to the division of space danger zone,
It is forbidden in real time/to prohibit the functions such as alarm out, camera shooting ball machine is tracked in conjunction with high speed, real-time linkage capture site operation video is captured and accused
Alert triggering photo, realizes the control of operation field solid.
Summary of the invention
The purpose of the present invention is to provide a kind of noninductive panorama cognitive methods of packaged type based on 3D laser, and this method can
Accomplish that remote centralized control is managed concentratedly, effectively mitigate the work load of monitoring personnel, improves working efficiency, greatly improve scene
Efficiency and field work safe are managed in operation.
To achieve the above object, the technical scheme is that a kind of noninductive panorama sense of packaged type based on 3D laser
Perception method, the collision detection algorithm of the laser beam and mobile object that are gone out by laser real time emission, with reflected laser beam
Area size identifies mobile object classification with contour feature;It then, is the contour feature of mobile personnel by mobile object classification
Independent identification to realize accurate spatial locations positioning of the mobile personnel in three-dimensional map scene, and then is realized and is strayed into, accidentally goes out
Regional Risk judgement, provides security alarm and prompt for supervisory control of substation person and on-site personnel.
In an embodiment of the present invention, the collision detection of the laser beam and mobile object that are gone out by laser real time emission is calculated
Method, with reflected laser beam area size, to identify mobile object classification, the specific implementation steps are as follows with contour feature:
Step S21, scanning in the space FOV is exported by 3D laser radar to turn the data of output to the location information data of point
It is changed to standard point cloud data format;
Step S22, several frame original point cloud datas are recorded, adds up and obtains background point cloud data;
Step S23, real-time point cloud data is obtained according to original point cloud data;
Step S24, real-time point cloud data and background dot cloud data comparison are obtained into difference point cloud data;
Step S25, discrepancy cloud data acquisition system is carried out clustering processing, obtains several clusters according to the distance between each point
The set of point, one mobile object of each set expression;
Step S26, according to contour feature, mobile object classification is identified.
It in an embodiment of the present invention, is the contour feature independent identification of mobile personnel by mobile object classification, thus real
The specific implementation steps are as follows for accurate spatial locations positioning of the existing mobile personnel in three-dimensional map scene:
Step S31,3D laser radar, which is matched by scanning element cloud with three-dimensional map, determines itself position, posture in the scene;
Step S32, it according to contour feature, identifies that mobile object classification is the set of the point of mobile personnel, represents and move from each
The point cloud data that a newest frame is taken out in the set of the point of moving member obtains the mobile personnel current location point cloud, and calculates and work as
The mass center of front position point cloud;
Step S33, using TOF method, the opposite 3D laser of mobile personnel is determined with the light velocity and transmitting receiving time difference product half
The distance of radar;
Step S34, mobile personnel position in laser coordinate system is transformed into scene coordinate system, realizes mobile personnel dimensionally
Accurate spatial locations positioning in figure field scape.
In an embodiment of the present invention, the mobile object classification includes mobile personnel, vehicle, bird.
Compared to the prior art, the invention has the following advantages:
1. the method for the present invention can make base's monitoring personnel become more familiar with field condition, and can accomplish remote centralized control
Centralized management effectively mitigates the work load of monitoring personnel, improves working efficiency;
2. the method for the present invention utilizes Airborne Lidar survey technology, the mobile objects such as substation operating personnel, vehicle are carried out accurate
Spatial perception greatly improves field operation control efficiency and field work safe;
3. the method for the present invention application high precision movement formula positioning device positions on-site personnel and manages, track record enriches
With improve substation's operation control technology, improve control efficiency;
4. the method for the present invention can be realized multi-platform, big data fusion, by greatly promote three-dimensional live platform analysis and early warning,
The ability of aid decision, final realize promote electric power netting safe running level, realize smart grid " flow of power, information flow, business
The height fusion and fine-grained management of stream ".
Detailed description of the invention
Fig. 1 is to realize mobile object classification method flow diagram by collision detection algorithm.
Fig. 2 is mobile personnel accurate spatial locations localization method flow chart.
Fig. 3 is to realize mobile personnel target identification schematic diagram by the method for the invention.
Fig. 4 is positioning schematic diagram of the mobile personnel in three-dimensional map scene.
Specific embodiment
With reference to the accompanying drawing, technical solution of the present invention is specifically described.
The present invention provides a kind of noninductive panorama cognitive methods of packaged type based on 3D laser, pass through laser real time emission
The collision detection algorithm of laser beam and mobile object out is identified with reflected laser beam area size with contour feature
Mobile object classification (including mobile personnel, vehicle, bird etc.);Then, mobile object classification is special for the profile of mobile personnel
Independent identification is levied, to realize accurate spatial locations positioning of the mobile personnel in three-dimensional map scene, and then realizes and is strayed into, accidentally
Regional Risk judges out, provides security alarm and prompt for supervisory control of substation person and on-site personnel.
As shown in Figure 1, the collision detection algorithm of the laser beam and mobile object gone out by laser real time emission, to be reflected back
To identify mobile object classification, the specific implementation steps are as follows with contour feature for the laser beam area size come:
Step S21, scanning in the space FOV is exported by 3D laser radar to turn the data of output to the location information data of point
It is changed to standard point cloud data format;
Step S22, several frame original point cloud datas are recorded, adds up and obtains background point cloud data;
Step S23, real-time point cloud data is obtained according to original point cloud data;
Step S24, real-time point cloud data and background dot cloud data comparison are obtained into difference point cloud data;
Step S25, discrepancy cloud data acquisition system is carried out clustering processing, obtains several clusters according to the distance between each point
The set of point, one mobile object of each set expression;
Step S26, according to contour feature, mobile object classification is identified.
As shown in Fig. 2, being the contour feature independent identification of mobile personnel by mobile object classification, to realize mobile personnel
The specific implementation steps are as follows for accurate spatial locations positioning in three-dimensional map scene:
Step S31,3D laser radar, which is matched by scanning element cloud with three-dimensional map, determines itself position, posture in the scene;
Step S32, it according to contour feature, identifies that mobile object classification is the set of the point of mobile personnel, represents and move from each
The point cloud data that a newest frame is taken out in the set of the point of moving member obtains the mobile personnel current location point cloud, and calculates and work as
The mass center of front position point cloud;
Step S33, using TOF method, the opposite 3D laser of mobile personnel is determined with the light velocity and transmitting receiving time difference product half
The distance of radar;
Step S34, mobile personnel position in laser coordinate system is transformed into scene coordinate system, realizes mobile personnel dimensionally
Accurate spatial locations positioning in figure field scape.
The following are specific implementation processes of the invention.
The present invention is based on the noninductive panorama cognitive methods of the packaged type of 3D laser, mainly pass through Study of Laser real time emission
The collision detection algorithm of laser beam and mobile personnel out is known by reflected laser beam area size with contour feature
It not to be not whether that personnel are mobile, and realize accurate spatial locations positioning of the mobile personnel in three-dimensional map scene, the method for the present invention
By laser sensor returned data, in conjunction with volume, movement speed, it is seen that it is mobile that light image can detect 30 meters of radius column regions
Personnel;It is implemented as follows:
1, as shown in Figure 1, capturing mobile personnel in laser point cloud scene:
Mobile personnel contour feature is mainly recognized by the laser beam that Study of Laser real time emission goes out, then by this feature
The algorithm that region independent identification comes out;
(1) location information that point is arrived in scanning in the space FOV is exported by three-dimensional laser sensor, by the data conversion of all outputs
For standard point cloud data format;
(2) several frame original point cloud datas are recorded, adds up and obtains a background point cloud data;
(3) real-time point cloud data is obtained according to original point cloud data;
(4) real-time point cloud data and background dot cloud data comparison are obtained into difference point cloud data;
(5) discrepancy cloud data acquisition system is carried out clustering processing, can obtains several cluster points according to the distance between each point
Set, one mobile object of each set expression.
(6) volume, movement speed are combined, it is seen that light camera picture captures mobile personnel.
By the above process, monitoring personnel is allow to see mobile personnel target in cloud scene, as shown in Figure 3.
2, mobile personnel positions in laser point cloud scene:
The algorithm of location information mainly in research point cloud scene where mobile personnel, passes through the information of Laser Discrete coordinate points
And Reflection intensity information, distance of the mobile personnel apart from laser radar apparatus is analyzed, to extrapolate mobile personnel in cloud
Coordinate information;
(1) laser radar, which is matched by scanning element cloud with three-dimensional map, determines itself position in the scene, posture.
(2) point cloud data for taking out a newest frame in closing is converged from each point for representing mobile object, can obtain the people
Member current location point cloud.Calculate the mass center of point cloud.
(3) 16 line three-dimensional lasers, distance test use TOF method, and the light velocity and transmitting receiving time difference product half determine
Mobile personnel relative laser distance.
(4) mobile personnel position in laser coordinate system is transformed into scene coordinate system
Through the above steps, the mobile personnel accurate location that can be will test out, and be accurately labeled in three-dimensional scenic;Make to supervise
Control personnel can see the location of mobile personnel in cloud scene, as being personnel position information in Fig. 4 box.
The above are preferred embodiments of the present invention, all any changes made according to the technical solution of the present invention, and generated function is made
When with range without departing from technical solution of the present invention, all belong to the scope of protection of the present invention.