CN115841467A - Method for identifying invasion foreign matters of power transmission line channel by fusing laser point cloud and visual image - Google Patents
Method for identifying invasion foreign matters of power transmission line channel by fusing laser point cloud and visual image Download PDFInfo
- Publication number
- CN115841467A CN115841467A CN202211517147.9A CN202211517147A CN115841467A CN 115841467 A CN115841467 A CN 115841467A CN 202211517147 A CN202211517147 A CN 202211517147A CN 115841467 A CN115841467 A CN 115841467A
- Authority
- CN
- China
- Prior art keywords
- point cloud
- transmission line
- power transmission
- cloud data
- line channel
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention discloses a method for identifying a foreign matter invaded by a power transmission line fusing laser and visible light.
Description
Technical Field
The invention relates to identification and positioning of an invasive foreign matter in a transmission line channel, realizes transmission line channel monitoring, and belongs to the technical field of electric power.
Background
At present, the operation and maintenance management mode of the power transmission line is to regularly send team personnel to inspect the power transmission line channel or rely on plane vision technology monitoring such as video/images and the like, the manual inspection has poor timeliness, low efficiency and high labor cost, and professional technicians are organized to detect and analyze when abnormality occurs in the power transmission line channel; the video/image isoplanar vision technology monitoring is greatly influenced by the environment, the night monitoring effect is poor, the shot visual image does not have space coordinates, the error is large during intelligent foreign matter identification and calibration in the later period, and a large amount of manual intervention is needed.
Along with the application of laser technology is more and more extensive, the monitoring of transmission line passageway has been realized through laser technology at present to appear, for example utilize unmanned aerial vehicle to carry on laser radar and patrol and examine the transmission line passageway, but laser monitoring also has misjudgement, misjudgement to and when the transmission line is thinner, laser probably has the shortcoming that can't scan the power line.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the method for identifying the invasion foreign matters of the power transmission line channel by fusing the laser point cloud and the visual image is provided, so that the defects of the prior art are overcome.
The technical scheme of the invention is as follows: a method for recognizing invasion foreign matters of a power transmission line channel by fusing laser point cloud and a visual image comprises the following steps:
step 1: collecting point cloud data and image data of the invading foreign matters in the power transmission line channel, establishing a spatial database, and generating a point cloud semantic segmentation model aiming at the point cloud data;
and 2, step: collectingInitial time t 0 Laser point cloud data and image data of the time-power transmission line channel, and the laser point cloud data and the image data are used as references, and t is assumed 0 No task foreign matter exists in the power transmission line channel at any moment;
and 3, step 3: point cloud segmentation, namely, placing the point cloud obtained in the step 2 into the model in the step 1 for segmentation, and segmenting a power line point cloud and a non-power line point cloud by adopting an automatic segmentation algorithm;
and 4, step 4: acquisition of t 1 、t 2 、...t n Laser point cloud data and image data of a power transmission line channel at a moment;
and 5: replacing the power line point cloud data segmented in the step 3 by equivalent, wherein the power line point cloud data in the power transmission line channel point cloud data acquired at different moments in the step 4 are replaced by the power line point cloud data segmented in the step 3;
step 6: setting a safety threshold, and generating n buffer space according to the safety threshold;
and 7: clipping t collected in step 4 with each buffer space n Checking the point cloud data at the moment to see whether the cutting result is empty, if the cutting result is empty, judging that no invasive foreign matter exists, and if the cutting result is not empty, indicating t n Invasion foreign matters possibly exist in the power transmission line channel at any moment;
and 8: judging the type of the invading foreign matter by utilizing the clipped point cloud data;
and step 9: and if the judgment can not be carried out in the step 8, fusing the clipped point cloud data and the image data, and carrying out calibration identification on the fused data.
Further, the point cloud semantic segmentation model is generated by training an invasive foreign matter point cloud sample data set by adopting a 3D point cloud semantic segmentation network RandLA Net.
Further, the step 6 specifically includes:
determining a safety threshold value d, selecting the power lines superposed in the step 5 one by one, generating a buffer space by taking the power lines as the center, d as the radius, L as the step length and the trend of the power lines as the advancing direction, and generating n buffer spaces L according to the length of the power lines 1 ，L 2 ，...L n 。
Further, the step 7 specifically includes:
clipping t collected in step 2 with each buffer space 0 The point cloud data of the moment and the cutting result are marked as P 1 、P 2 、...P n Let t be 0 If no potential safety hazard exists in the power transmission line channel at the initial moment, no point cloud data, namely P, exists in the safety space 1 ＝P 2 ＝...＝P n = null; clipping the t collected in step 5 with each buffer space as well n The point cloud data of the moment and the cutting result are recorded as Q 1 、Q 2 、...Q n Viewing the cutting result Q n Whether or not it is empty, if Q n If not, = null, then t can be determined n The state t of the time of day transmission line channel 0 The time is consistent, the invasion foreign matter is not existed, if Q n Not equal to null, i.e. Q n ≠P n Description of t n There may be an intruding foreign object in the transmission line channel at the moment.
Further, the step 8 specifically includes: if Q n ≠P n Then assume that L is used 1 Cutting t n Point cloud data collected at any moment and Q for cutting result n1 Is shown if Q n1 ≠P 1 And Q n1 Not equal to null, then L of the power line 1 Segment at t n The potential safety hazard exists at any moment, and then Q is added n1 Implanting the point cloud data into the model in the step 2, dividing the point cloud data into Q n1 Intrinsic t n What category the point cloud data of the moment belongs to, if Q n1 The existing point cloud data is a whole, and the foreign matters invaded by the power transmission line can be directly judged after segmentation.
Further, the step 9 specifically includes: if Q is in step 8 n1 If the point cloud data existing in the device is not complete, the t acquired by the device in the step 2 is used n The image of the power transmission line channel at the moment is recorded as M n A channel t of a power transmission line n Point cloud of time Q n And image M n Performing fusion, and marking t in fused data n The invasion of foreign matters at all times.
Further, the step 9 further includes: and measuring the distance between the foreign matter and the power line through the point cloud data.
The method for identifying the intrusion foreign matter in the power transmission line channel by fusing the laser point cloud and the visual image according to claim 5, wherein the safety threshold d is determined by standards GB50545 and DL/T741.
The invention has the beneficial effects that: compared with the prior art, the method combines laser and visible light, collects point cloud and visual image data in the power transmission line channel through laser and visible light integrated monitoring equipment, firstly extracts foreign matters by using the point cloud data, and fuses the visual image data and the point cloud data to calibrate the invading foreign matters when the point cloud data is insufficient to identify and judge the foreign matters, thereby improving the accuracy of identifying and positioning the invading foreign matters.
Drawings
FIG. 1 is a schematic view of a buffer space according to the present invention;
FIG. 2 is a schematic view of the present invention showing the intrusion foreign object entirely contained within the buffer space;
FIG. 3 is a schematic diagram of the invention showing only a portion of the structure of an intruding foreign object being contained in the buffer space;
FIG. 4 is a flow chart of the present invention.
Detailed Description
In order to better understand the technical solutions, the technical solutions will be described in detail below with reference to the drawings and the detailed description.
Example 1 was carried out: referring to fig. 4, the method for identifying the intrusion foreign object in the power transmission line channel by fusing the laser point cloud and the visual image includes:
before the invasion foreign body is identified, preparation work is required: laser and visible light integrated equipment is installed, and the equipment is installed on an iron tower and faces to the direction of a power transmission line channel;
step 1: collecting point cloud data and image data of the invading foreign matters in a power transmission line channel, establishing a spatial database, training a point cloud sample data set of the invading foreign matters by adopting a lightweight high-efficiency large-scale 3D point cloud semantic segmentation network RandlaNet aiming at the point cloud data, and generating a point cloud semantic segmentation model;
step 2: collecting an initial time t 0 Laser point cloud data and image data of the time-power transmission line channel, and the laser point cloud data and the image data are used as references, and t is assumed 0 No task foreign matter exists in the power transmission line channel at any moment;
and 3, step 3: point cloud segmentation: placing the point cloud obtained in the step 2 into the model in the step 1 for segmentation, and segmenting a power line point cloud and a non-power line point cloud by adopting an automatic segmentation algorithm; putting the point cloud data acquired in the step 2 into the algorithm model in the step 1, and segmenting different types of point cloud data, such as vegetation, roads, towers, wires, buildings and the like;
and 4, step 4: acquisition of t 1 、t 2 、...t n Laser point cloud data and image data of a power transmission line channel at a moment;
and 5: replacing the power line point cloud data segmented in the step 3 by equivalent, wherein the power line point cloud data in the power transmission line channel point cloud data acquired at different moments in the step 4 are replaced by the power line point cloud data segmented in the step 3; because the point cloud data acquired in the step 3 and the step 4 are in the same coordinate system, the power line point cloud in the step 3 can be superposed into the point cloud in the step 4;
step 6: setting a safety threshold, determining a safety threshold d according to GB50545 (110 kV-750 kV overhead transmission line design specification) and DL/T741 (overhead transmission line operation specification), selecting power lines superposed in the step 5 one by one, taking the power lines as centers, d as radius, L as step length and the power line direction as advancing direction, generating a buffer space, and generating n buffer spaces according to the length of the power lines, wherein L is L 1 ，L 2 ，...L n The space of the buffer zone is shown in figure 1;
and 7: clipping t collected in step 4 with each buffer space n The point cloud data of the moment and the cutting result are marked as P 1 、P 2 、...P n Let t be 0 No potential safety hazard exists in the power transmission line channel at the initial moment, and no any potential safety hazard exists in the safety spacePoint cloud, i.e. P 1 ＝P 2 ＝...＝P n = null; clipping the t collected in step 5 with each buffer space as well n The point cloud data of the moment and the cutting result are recorded as Q 1 、Q 2 、...Q n Looking over the cutting result Q n Whether or not it is empty, if Q n If not, = null, then t can be determined n The state t of the time of day transmission line channel 0 The time is consistent, the invasion foreign matter is not existed, if Q n Not equal to null, i.e. Q n ≠P n Description of t n Step 8, if the power transmission line channel possibly has invasion foreign matters at all times;
and step 8: suppose with L 1 Cutting t n Point clouds collected at all times, and cutting results Q n1 Is shown if Q n1 ≠P 1 And Q n1 Not equal to null, L for power line is illustrated 1 Segment at t n The potential safety hazard exists all the time, namely, in the safe distance d around the power line, the point cloud of other objects exists, and at the moment, Q is added n1 Implanting the point cloud data into the model in the step 2, dividing the point cloud data into Q n1 Intrinsic t n What category the point cloud of the moment belongs to, if Q n1 Existing point clouds are a whole (such as a complete kite outline), and as shown in fig. 2, foreign matters invading the power transmission line can be directly judged after segmentation; if Q is n1 The existing point cloud is not complete (for example, only a certain part of the kite is cut, if the kite is just near the buffer space, only a part of the kite point cloud can be cut), as shown in fig. 3, then step 9 is performed;
and step 9: to distinguish Q n1 The point cloud data with incomplete internal structure is the invasive foreign matter, and t acquired by the equipment in the step 3 is combined n The image of the power transmission line channel at the moment is recorded as M n Passing the power transmission line through a channel t n Point cloud of time Q n And image M n Performing fusion, and marking t in fused data n And (3) constantly invading the foreign matters, and measuring the distance between the foreign matters and the power line through point cloud.
The foregoing is a more detailed description of the invention in connection with specific preferred embodiments and it is not intended that the invention be limited to these specific details. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all shall be considered as belonging to the protection scope of the invention.
Claims (8)
1. A method for recognizing invasion foreign matters of a power transmission line channel by fusing laser point cloud and a visual image is characterized by comprising the following steps:
step 1: collecting point cloud data and image data of the invading foreign matters in the power transmission line channel, establishing a spatial database, and generating a point cloud semantic segmentation model aiming at the point cloud data;
step 2: collecting an initial time t 0 Laser point cloud data and image data of the time-power transmission line channel, and the laser point cloud data and the image data are used as references, and t is assumed 0 No task foreign matter exists in the power transmission line channel at any moment;
and step 3: point cloud segmentation, namely putting the point cloud obtained in the step 2 into the model in the step 1 for segmentation, and segmenting a power line point cloud and a non-power line point cloud by adopting an automatic segmentation algorithm;
and 4, step 4: acquisition of t 1 、t 2 、...t n Laser point cloud data and image data of a power transmission line channel at a moment;
and 5: replacing the power line point cloud data segmented in the step 3 by equivalent, wherein the power line point cloud data in the power transmission line channel point cloud data acquired at different moments in the step 4 are replaced by the power line point cloud data segmented in the step 3;
step 6: setting a safety threshold, and generating n buffer space according to the safety threshold;
and 7: clipping t collected in step 4 with each buffer space n Checking the point cloud data at the moment to see whether the cutting result is empty, if the cutting result is empty, judging that no invasive foreign matter exists, and if the cutting result is not empty, indicating t n Invasion foreign matters possibly exist in the power transmission line channel at any moment;
and 8: judging the type of the invading foreign matter by utilizing the clipped point cloud data;
and step 9: and if the judgment can not be carried out in the step 8, fusing the clipped point cloud data and the image data, and carrying out calibration identification on the fused data.
2. The method for identifying the intrusion foreign matter in the power transmission line channel by fusing the laser point cloud and the visual image according to claim 1, wherein the point cloud semantic segmentation model is generated by training an intrusion foreign matter point cloud sample data set by a 3D point cloud semantic segmentation network RandlANT.
3. The method for identifying the intrusion foreign matter in the power transmission line channel by fusing the laser point cloud and the visual image according to claim 1, wherein the step 6 specifically comprises the following steps:
determining a safety threshold value d, selecting the power lines superposed in the step 5 one by one, generating a buffer space by taking the power lines as the center, d as the radius, L as the step length and the trend of the power lines as the advancing direction, and generating n buffer spaces L according to the length of the power lines 1 ，L 2 ，...L n 。
4. The method for identifying the intrusion foreign matter in the power transmission line channel by fusing the laser point cloud and the visual image according to claim 1, wherein the step 7 specifically comprises the following steps:
clipping t collected in step 2 with each buffer space 0 The point cloud data of the moment and the cutting result are marked as P 1 、P 2 、...P n Let t be 0 If no potential safety hazard exists in the power transmission line channel at the initial moment, no point cloud data, namely P, exists in the safety space 1 ＝P 2 ＝...＝P n = null; clipping the t collected in step 5 with each buffer space as well n The point cloud data of the moment and the cutting result are recorded as Q 1 、Q 2 、...Q n Viewing the cutting result Q n Whether or not it is empty, if Q n If not, = null, then t can be determined n State of time of day transmission line path and t 0 Coincide in time, sayNo invasion of foreign body, if Q n Not equal to null, i.e. Q n ≠P n Description of t n There may be intruding foreign objects in the transmission line channel at all times.
5. The method for identifying the intrusion foreign matter in the power transmission line channel by fusing the laser point cloud and the visual image according to claim 4, wherein the step 8 specifically comprises the following steps: if Q is n ≠P n Then assume that L is used 1 Cutting t n Point cloud data collected at any moment, Q for cutting result n1 Is shown if Q n1 ≠P 1 And Q n1 Not equal to null, then L of the power line 1 Segment at t n The potential safety hazard exists at any moment, and then Q is added n1 Implanting the point cloud data into the model in the step 2, dividing the point cloud data into Q n1 Intrinsic t n What category the point cloud data of the moment belongs to, if Q n1 The existing point cloud data is a whole, and the foreign matters invaded by the power transmission line can be directly judged after segmentation.
6. The method for identifying the intrusion foreign matter in the power transmission line channel by fusing the laser point cloud and the visual image according to claim 5, wherein the step 9 specifically comprises: if Q is in step 8 n1 If the point cloud data in the point cloud data is incomplete, the t acquired by the equipment in the step 2 is used n The image of the power transmission line channel at the moment is recorded as M n Passing the power transmission line through a channel t n Point cloud of time Q n And image M n Performing fusion, and marking t in fused data n The invasion of foreign matters at all times.
7. The method for identifying the intrusion foreign matter in the power transmission line channel by fusing the laser point cloud and the visual image according to claim 5, wherein the step 9 further comprises: and measuring the distance between the foreign matter and the power line through the point cloud data.
8. The method for identifying the intrusion foreign matter in the power transmission line channel by fusing the laser point cloud and the visual image according to claim 5, wherein the safety threshold d is determined by standards GB50545 and DL/T741.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211517147.9A CN115841467A (en) | 2022-11-29 | 2022-11-29 | Method for identifying invasion foreign matters of power transmission line channel by fusing laser point cloud and visual image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211517147.9A CN115841467A (en) | 2022-11-29 | 2022-11-29 | Method for identifying invasion foreign matters of power transmission line channel by fusing laser point cloud and visual image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115841467A true CN115841467A (en) | 2023-03-24 |
Family
ID=85576269
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211517147.9A Pending CN115841467A (en) | 2022-11-29 | 2022-11-29 | Method for identifying invasion foreign matters of power transmission line channel by fusing laser point cloud and visual image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115841467A (en) |
-
2022
- 2022-11-29 CN CN202211517147.9A patent/CN115841467A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112633535A (en) | Photovoltaic power station intelligent inspection method and system based on unmanned aerial vehicle image | |
| CN108957240A (en) | Electric network fault is remotely located method and system | |
| CN108037770A (en) | Unmanned plane power transmission line polling system and method based on artificial intelligence | |
| CN108680833B (en) | Composite insulator defect detection system based on unmanned aerial vehicle | |
| CN101625723A (en) | Rapid image-recognizing method of power line profile | |
| CN112684806A (en) | Electric power inspection unmanned aerial vehicle system based on dual obstacle avoidance and intelligent identification | |
| CN117214617A (en) | Smart power grid fault real-time monitoring and positioning system and method | |
| CN117079397B (en) | Wild human and animal safety early warning method based on video monitoring | |
| CN111507189B (en) | Insulator string defect rapid detection method based on image processing technology | |
| CN116258980A (en) | Unmanned aerial vehicle distributed photovoltaic power station inspection method based on vision | |
| CN111244822A (en) | Fixed-wing unmanned aerial vehicle line patrol method, system and device in complex geographic environment | |
| CN113066070A (en) | Multi-source data fusion interaction method in three-dimensional scene | |
| Li et al. | The future application of transmission line automatic monitoring and deep learning technology based on vision | |
| CN113722870B (en) | Power grid operation full-space monitoring method, system, equipment and medium | |
| CN115664006A (en) | Increment distribution network intelligence management and control integration platform | |
| CN109697410B (en) | Remote sensing ground object identification method for overhead transmission line coverage area | |
| CN114445782A (en) | Power transmission line image acquisition system based on edge AI and Beidou short messages | |
| CN117791864A (en) | Remote inspection system for power line | |
| CN106155091A (en) | A kind of unmanned plane barrier-avoiding method and device | |
| CN109712018A (en) | A kind of power transmission line comprehensive monitoring system | |
| CN113780237A (en) | External damage prevention early warning method, device and system for underground pipeline | |
| CN115841467A (en) | Method for identifying invasion foreign matters of power transmission line channel by fusing laser point cloud and visual image | |
| CN112801432A (en) | Fan unit blade intelligent inspection system and fan unit blade inspection method | |
| CN116363397A (en) | Equipment fault checking method, device and inspection system | |
| CN113724247B (en) | Intelligent substation inspection method based on image discrimination technology |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |