CN-121978667-A - Depth vision-based transmission line corridor monocular image ranging method and system

CN121978667ACN 121978667 ACN121978667 ACN 121978667ACN-121978667-A

Abstract

The invention discloses a depth vision-based power transmission line corridor monocular image ranging method and system, and relates to the technical field of image ranging; the method comprises the steps of importing three-dimensional point cloud data of a power transmission line corridor into edge computing equipment, establishing a mapping relation matrix of a pixel coordinate system and a three-dimensional point cloud space coordinate system, positioning space coordinate information corresponding to a target dangerous object in the three-dimensional point cloud space, determining real-time space distance between the target dangerous object and the power transmission line, and generating a grading alarm signal. The method solves the technical problems that dangerous objects in the corridor of the power transmission line cannot be accurately identified in real time in the prior art, a dynamic distance monitoring and grading alarm mechanism is lacked, and the safety of power transmission is difficult to ensure, and achieves the technical effects of real-time monitoring and grading alarm of dangerous objects in the corridor of the power transmission line, and improving the power inspection efficiency and safety.

Inventors

LI XUEJUN
WANG XIAOPENG
DAI XIANGLONG
JIANG YONG
HE CHENGHU

Assignees

江苏濠汉信息技术有限公司

Dates

Publication Date: 20260505
Application Date: 20260408

Claims (10)

1. The utility model provides a transmission line corridor monocular image ranging method based on depth vision which is characterized in that the method comprises the following steps: acquiring two-dimensional image information of a corridor area of a power transmission line by using monocular image acquisition equipment deployed on the power transmission tower, constructing a lightweight target detection model facing edge computing equipment, carrying out hazard source identification on the two-dimensional image information, and outputting a two-dimensional pixel coordinate set with a hazard source category mark; Importing three-dimensional point cloud data of a power transmission line corridor into the edge computing equipment, and establishing a mapping relation matrix of a pixel coordinate system and a three-dimensional point cloud space coordinate system; Using the mapping relation matrix, positioning space coordinate information corresponding to a target dangerous object in a three-dimensional point cloud space through a two-dimensional pixel coordinate set with a dangerous source category mark, wherein the space coordinate information is mapped in association with each element in the two-dimensional pixel coordinate set; based on the space coordinate information, determining a real-time space distance between the target dangerous object and the power transmission line, and judging whether the real-time space distance is smaller than a preset safety threshold value or not; and if the judgment result is yes, triggering a local early warning mechanism to generate a grading warning signal.
2. The depth vision-based transmission line corridor monocular image ranging method of claim 1, wherein a mapping relation matrix of a pixel coordinate system and a three-dimensional point cloud space coordinate system is established, the method further comprising: Correcting distortion errors by utilizing the spatial distribution characteristics of the transmission wires in the three-dimensional point cloud data of the transmission line corridor according to the camera focal length and main point offset parameters of the monocular image acquisition equipment and combining the camera focal length and main point offset parameters; Meanwhile, M uniformly distributed transmission tower characteristic points are selected from the three-dimensional point cloud data of the transmission line corridor, first coordinate information in the three-dimensional point cloud space coordinate system and second coordinate information in the pixel coordinate system are obtained, and a nonlinear least square equation set is constructed, wherein M is more than or equal to 6; and performing projection error control according to the nonlinear least square equation set.
3. The depth vision-based transmission line corridor monocular image ranging method of claim 2, wherein the method further comprises: based on the monocular visual image stream acquired by the monocular image acquisition equipment, carrying out real-time defogging treatment, and carrying out super-resolution reconstruction on the identified dangerous source area; And meanwhile, eliminating coordinate jump errors caused by jitter through time sequence filtering to obtain the two-dimensional image information.
4. A depth vision based transmission line corridor monocular image ranging method as claimed in claim 3, wherein in combination with the camera focal length, principal point offset parameters, distortion errors are corrected, the method further comprising: according to the installation position parameters of the transmission tower, establishing a rigid transformation relation from a pixel coordinate system to a transmission line corridor area; and correcting the spatial distribution characteristics of the power transmission wires according to the rigid transformation relation from the pixel coordinate system to the corridor area of the power transmission line.
5. The depth vision based transmission line corridor monocular image ranging method of claim 4, wherein the spatial distribution characteristics of the transmission line are corrected, before the method further comprises: acquiring original point cloud data of a power transmission corridor; filtering the original point cloud data, wherein the filtering is used for removing vegetation interference points and topography interference points; and dividing the filtered point cloud data into a plurality of independent blocks according to the transmission tower position, and converting the independent blocks into an encryption compression format which can be analyzed by the edge computing equipment.
6. The depth vision-based transmission line corridor monocular image ranging method of claim 1, wherein a real-time spatial distance between a target hazard and a transmission line is determined based on the spatial coordinate information, the method comprising: taking Euclidean distance between a target dangerous object and a power transmission line as the real-time space distance; Meanwhile, lead swing parameters are introduced, dynamic compensation is carried out on the Euclidean distance, and the compensation quantity is dynamically adjusted according to real-time wind speed data.
7. The depth vision-based transmission line corridor monocular image ranging method of claim 1, wherein a lightweight target detection model facing an edge computing device is constructed, the method comprising: adopting MobileNetV3 as a characteristic to extract a main network, and adding a space attention unit after the last layer of convolution layer; and based on the feature extraction backbone network, migrating model parameters obtained by pre-training to a lightweight model, and carrying out model quantification when the edge computing equipment is deployed by using a spatial attention unit.
8. The depth vision-based transmission line corridor monocular image ranging method of claim 7, wherein the method comprises: Establishing a shared memory channel between a first chip associated with the edge computing device and a second chip associated with the lightweight target detection model, and determining a first ranging delay factor; and dynamically distributing computing resources of the first chip and the second chip based on the task priority, determining a second ranging delay factor, and combining the first ranging delay factor to perform ranging task response optimization processing under delay compensation.
9. A depth vision based transmission line corridor monocular image ranging method as claimed in claim 8, and converted into an encrypted compressed format resolvable by the edge computing device, the method comprising: When the edge computing equipment loads point cloud data in an encryption compression format, hash check is used for comparing topological structure characteristics of an independent block corresponding to the point cloud data and M adjacent independent blocks, wherein M is more than or equal to 3; if the deviation value of the hash value of the consistency of the spatial curvature distribution and the normal vector and the pre-stored reference value is smaller than the preset deviation value, continuing to execute the decompression operation of the point cloud data.
10. A depth vision based transmission line corridor monocular image ranging system for implementing the depth vision based transmission line corridor monocular image ranging method of any one of claims 1 to 9, the system comprising: The dangerous source identification module is used for acquiring two-dimensional image information of a corridor area of the power transmission line by using monocular image acquisition equipment deployed on the power transmission tower, constructing a lightweight target detection model facing edge computing equipment, carrying out dangerous source identification on the two-dimensional image information, and outputting a two-dimensional pixel coordinate set with a dangerous source category mark; The relation matrix establishing module is used for importing three-dimensional point cloud data of the power transmission line corridor into the edge computing equipment and establishing a mapping relation matrix of a pixel coordinate system and a three-dimensional point cloud space coordinate system; The space coordinate information positioning module is used for positioning space coordinate information corresponding to a target dangerous object in a three-dimensional point cloud space through a two-dimensional pixel coordinate set with a dangerous source category mark by using the mapping relation matrix, and the space coordinate information is mapped in an associated mode with each element in the two-dimensional pixel coordinate set; The space distance determining module is used for determining the real-time space distance between the target dangerous object and the power transmission line based on the space coordinate information and judging whether the real-time space distance is smaller than a preset safety threshold value or not; and the hierarchical alarm signal generation module is used for triggering the local early warning mechanism to generate the hierarchical alarm signal if the judgment result is yes.

Description

Depth vision-based transmission line corridor monocular image ranging method and system Technical Field The invention relates to the technical field of image ranging, in particular to a power transmission line corridor monocular image ranging method and system based on depth vision. Background With the promotion of smart grid construction, the safety monitoring of a power transmission line corridor faces higher requirements. The traditional inspection method has the problems of low efficiency, high cost, limitation of terrain and climate and the like, is difficult to realize all-weather and full-area coverage, and the traditional automatic monitoring technology relies on complex deep learning models, has the bottlenecks of insufficient calculation power, delayed response and the like when being deployed on edge computing equipment, and cannot meet the real-time detection requirement of dangerous objects. Meanwhile, part of schemes can only identify target categories, lack accurate quantification of space distance between dangerous objects and a power transmission line, and are difficult to trigger classified warning by combining a safety threshold. The prior art has the technical problems that dangerous objects in a power transmission line corridor cannot be accurately identified in real time, a dynamic distance monitoring and grading alarm mechanism is lacked, and power transmission safety is difficult to guarantee. Disclosure of Invention The application provides a depth vision-based power transmission line corridor monocular image ranging method and system, which are used for solving the technical problems that in the prior art, dangerous objects in a power transmission line corridor cannot be accurately identified in real time, a dynamic distance monitoring and grading alarm mechanism is lacking, and the safety of power transmission is difficult to guarantee. In view of the above problems, the application provides a depth vision-based transmission line corridor monocular image ranging method and system. In a first aspect of the present application, there is provided a depth vision-based transmission line corridor monocular image ranging method, the method comprising: The method comprises the steps of acquiring two-dimensional image information of a corridor area of a power transmission line by using monocular image acquisition equipment deployed by a power transmission tower, constructing a lightweight target detection model facing edge computing equipment, carrying out hazard source identification on the two-dimensional image information, outputting a two-dimensional pixel coordinate set with a hazard source category mark, importing three-dimensional point cloud data of the power transmission line into the edge computing equipment, establishing a mapping relation matrix of a pixel coordinate system and a three-dimensional point cloud space coordinate system, locating space coordinate information corresponding to a target hazard object in the three-dimensional point cloud space by using the mapping relation matrix through the two-dimensional pixel coordinate set with the hazard source category mark, determining real-time space distance between the target hazard object and the power transmission line based on the space coordinate information, judging whether the real-time space distance is smaller than a preset safety threshold, and triggering a local early warning mechanism to generate a grading warning signal if the real-time distance is smaller than the preset safety threshold. In a second aspect of the present application, there is provided a depth vision based transmission line corridor monocular image ranging system, the system comprising: The system comprises a dangerous source identification module, a relation matrix establishing module, a space coordinate information positioning module and a classified warning signal generating module, wherein the dangerous source identification module is used for acquiring two-dimensional image information of a transmission line corridor area by using a monocular image acquisition device deployed by a transmission tower, constructing a lightweight target detection model facing an edge computing device, carrying out dangerous source identification on the two-dimensional image information, outputting a two-dimensional pixel coordinate set with a dangerous source type mark, the relation matrix establishing module is used for guiding transmission line corridor three-dimensional point cloud data into the edge computing device, establishing a mapping relation matrix of a pixel coordinate system and a three-dimensional point cloud space coordinate system, the space coordinate information positioning module is used for using the mapping relation matrix, positioning space coordinate information corresponding to a target dangerous object in a three-dimensional point cloud space by the two-dimensional pixel coordinate set with the dangerous source type mark, the space coo