Transmission line distance measurement method, system, equipment and medium based on drone images

Abstract

The present invention relates to the technical field of power transmission line monitoring, and in particular to a power transmission line ranging method, system and medium based on drone images. In the ranging method, by extracting features of a target line, features of the target power transmission line are obtained; based on a binocular image acquisition device, images of a target area are acquired to obtain a target area environment video; based on the features of the target power transmission line, target line positioning is performed in an image frame to obtain target line positioning information; based on the target line positioning information, target line positioning matching is performed to obtain target line matching information; stereo matching analysis is performed according to the target line positioning information, the target line matching information and the ROI area position to obtain the target distance, so as to accurately know whether there are environmental obstacles that interfere with the normal power transmission operation of the line in the operating environment where the target line is located, and improve the operation and maintenance efficiency and timeliness of the removal of transmission line obstacles.

Description

Technical Field

[0001] The present invention relates to the field of power transmission line monitoring technology, and in particular to a power transmission line distance measurement method, system, equipment and medium based on drone images. Background Art

[0002] The power industry is constantly developing and innovating with the needs of residents and social industrial development. The scale of transmission line operation for electric energy transmission has grown and become increasingly prosperous, and the operating environment of transmission lines has become increasingly complex. Transmission lines are often shut down due to collisions with surrounding buildings, displacement and deformation caused by the oblique branches of trees, and entanglement with artificial foreign objects such as plastic film in greenhouses.

[0003] At present, the methods to prevent foreign objects from entangled in transmission lines are relatively crude. For example, video surveillance devices are installed on transmission line supporting towers and poles, and obstacles that interfere with the safe operation of transmission lines are cleared and removed based on manual judgment. While consuming manpower and material resources, the accuracy and effectiveness of eliminating obstacles on transmission lines are also low.

[0004] In summary, the prior art has a technical problem that the spatial distance between obstacles and transmission lines cannot be accurately measured, resulting in obstacles interfering with the safe operation of transmission lines. Summary of the invention

[0005] In response to the problems in the prior art, the present invention provides a transmission line distance measurement method, system, equipment and medium based on drone images, which can quickly measure the spatial distance between the transmission line and environmental obstacles, and improve the operation and maintenance efficiency and timeliness of transmission line obstacle removal.

[0006] In order to solve the above technical problems, the technical solution of the present invention is:

[0007] In a first aspect, a transmission line distance measurement method based on drone images is provided, comprising:

[0008] Extract features of the target line to obtain the features of the target transmission line;

[0009] Based on the binocular image acquisition device, the target area image is acquired to obtain the target area environment video, wherein the target area environment video includes the first environment video information and the second environment video information;

[0010] Acquire an image frame based on the first environmental video information;

[0011] Determining whether the image frame is an initial frame;

[0012] If the image frame is an initial frame, locating the target line in the image frame based on the target power transmission line feature to obtain target line location information;

[0013] Extracting the ROI area based on the target line positioning information;

[0014] Performing regional positioning matching based on the ROI region in the second environment video information to obtain ROI region matching information;

[0015] A stereo matching analysis is performed according to the target line positioning information, the ROI area and the ROI area matching information to obtain a target distance set.

[0016] In the second aspect, a transmission line distance measurement system based on drone images is provided, including: a feature extraction execution module, an image acquisition execution module, an image frame acquisition module, an image frame judgment module, a line feature positioning module, an interest area extraction module, an area matching execution module, and a target distance calculation module, wherein:

[0017] A feature extraction execution module is used to extract features of a target line and obtain features of a target transmission line;

[0018] An image acquisition execution module is used to acquire an image of a target area based on a binocular image acquisition device to obtain an environment video of the target area, wherein the environment video of the target area includes first environment video information and second environment video information;

[0019] An image frame acquisition module, used to acquire an image frame based on the first environment video information;

[0020] An image frame determination module, used to determine whether the image frame is an initial frame;

[0021] A line feature positioning module, configured to locate the target line in the image frame based on the target power transmission line feature if the image frame is an initial frame, and obtain target line positioning information;

[0022] A region of interest extraction module, used to extract a ROI region based on the target line positioning information;

[0023] A region matching execution module, used for performing region positioning matching based on the ROI region in the second environment video information to obtain ROI region matching information;

[0024] The target distance calculation module is used to perform stereo matching analysis according to the target line positioning information, the ROI area and the ROI area matching information to obtain a target distance set.

[0025] According to a third aspect, a computer device is provided, comprising:

[0026] Processor; and

[0027] A memory, configured to store executable instructions of the processor;

[0028] Wherein, the processor is configured to execute the above-mentioned transmission line ranging method based on drone images by executing the executable instructions.

[0029] In a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned transmission line ranging method based on drone images is implemented.

[0030] The present invention has the following beneficial effects:

[0031] The present invention solves the technical problem in the prior art that the spatial distance between obstacles and transmission lines cannot be accurately measured, resulting in obstacles interfering with the safe operation of transmission lines, and achieves the technical effect of quickly measuring the spatial distance between transmission lines and environmental obstacles and improving the operation and maintenance efficiency and timeliness of transmission line obstacle removal. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a flow chart of the transmission line distance measurement method based on UAV images of the present invention;

[0033] Figure 2 The flow chart of obtaining the inspection cycle of the target line also included in step S300 of the distance measurement method of the present invention;

[0034] Figure 3 A flowchart of obtaining a primary inspection cycle in step S330 of the distance measurement method of the present invention;

[0035] Figure 4 This is an operation flow chart of step S500 of the distance measurement method of the present invention when the image frame is not an initial frame;

[0036] Figure 5 Flow chart of obtaining a target distance set in step S800 of the distance measurement method of the present invention;

[0037] Figure 6 This is a flow chart of obtaining a three-dimensional scene in the ROI area in step S830 of the distance measurement method of the present invention;

[0038] Figure 7 Flow chart of obtaining a target distance set in step S860 of the distance measurement method of the present invention;

[0039] Figure 8 The present invention is a schematic structural block diagram of a power transmission line distance measurement system based on UAV images.

[0040] Fig. 9The present invention is a block diagram of a computer device according to an embodiment of the present invention. Implementation

[0041] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0042] Figure 1 A flow chart of a transmission line distance measurement method based on drone images in an embodiment of the present disclosure is shown.

[0043] refer to Figure 1 The present invention provides a transmission line distance measurement method based on drone images, which is applied to a transmission line distance measurement system. The transmission line distance measurement system is communicatively connected to a binocular image acquisition device, and includes:

[0044] S100: extracting features of the target line to obtain features of the target transmission line;

[0045] S200: performing target area image acquisition based on the binocular image acquisition device to obtain a target area environment video, wherein the target area environment video includes first environment video information and second environment video information;

[0046] S300: Acquire an image frame based on the first environment video information;

[0047] S400: Determine whether the image frame is an initial frame;

[0048] S500: If the image frame is an initial frame, locate the target line in the image frame based on the target power transmission line feature to obtain target line location information;

[0049] S600: extracting a ROI region based on the target line positioning information;

[0050] S700: Performing regional positioning matching based on the ROI region in the second environment video information to obtain ROI region matching information;

[0051] S800: Perform stereo matching analysis according to the target line positioning information, the ROI area and the ROI area matching information to obtain a target distance set.

[0052] Specifically, by extracting features of the target line, the features of the target transmission line are obtained, and the features of the target transmission line provide feature references for subsequent target line identification in the video image; based on the binocular image acquisition device, target area image acquisition is performed to obtain a target area environment video, wherein the target area environment video includes first environment video information and second environment video information; an image frame is acquired based on the first environment video information; it is determined whether the image frame is an initial frame; if the image frame is an initial frame, the target line is located in the image frame based on the features of the target transmission line to obtain target line location information; an ROI area is extracted based on the target line location information, and a target area is obtained by performing a scan of the image frame to obtain the target area location information; The target line is divided into a region of interest (ROI) in the target line to reduce the amount of data compared with the second video information in the subsequent distance measurement; the target line is positioned and matched based on the ROI area in the second environmental video information to obtain the target line matching information; a stereo matching analysis is performed according to the target line positioning information, the target line matching information and the ROI area position to obtain the target distance, so as to accurately know whether there are environmental obstacles that interfere with the normal power transmission operation of the line in the operating environment of the target line, thereby assisting the operation and maintenance management personnel to perform the maintenance and management of the transmission line, and achieving the technical effect of quickly measuring the spatial distance between the transmission line and the environmental obstacle, and improving the operation and maintenance efficiency and timeliness of the transmission line obstacle removal.

[0053] The following are Figure 1 Each step in is described in detail.

[0054] In step S100, feature extraction is performed on the target line to obtain the target power transmission line features.

[0055] Specifically, the target line is a transmission line erected on buildings such as transmission towers and cement poles for transmitting electric energy. To ensure the safe operation of the line, the power department needs to inspect the line to prevent intruders from approaching or entangled in the transmission line and causing power transmission safety accidents.

[0056] In this embodiment, based on the production characteristics of uniformity of transmission line specifications and the normative characteristics of transmission line erection, the target line in power transmission operation is imaged, and image features are extracted based on the target line image acquisition results to obtain the target transmission line features of the target line, the target transmission line features include color features and appearance structure features, for example, the color feature is RGB value (41, 36, 33), and the appearance structure feature is three-phase four-wire parallel layout. The target transmission line features provide feature references for subsequent target line identification in video images.

[0057] In step S200, target area image acquisition is performed based on the binocular image acquisition device to obtain a target area environment video, wherein the target area environment video includes first environment video information and second environment video information.

[0058] Specifically, in this embodiment, the binocular image acquisition device is a binocular vision lens mounted on a drone. The binocular image acquisition device is composed of two left and right cameras with parallel optical axes. The drone is controlled to perform video image acquisition on the target line installation environment in a certain area. For example, video image acquisition is performed on power transmission line installation and operation scenes such as steep mountains and complex road intersections to obtain the target area environment video. The target area environment video is used for subsequent analysis to determine whether there are tree obstacles and other risk factors that affect the safe operation of the transmission line in the transmission line installation environment.

[0059] The first environmental video information is acquired based on the left / right camera, and the second environmental video information is acquired based on the right / left camera. Both the first environmental video information and the second environmental video information are two-dimensional running image data obtained by synchronous image acquisition of the same frame number of each target object in the same scene.

[0060] In step S300, an image frame is acquired based on the first environmental video information.

[0061] like Figure 2 As shown, step S300 of the method provided by the present invention also includes:

[0062] S310: Extracting building features based on the image frame to obtain environmental building information;

[0063] S320: Extracting traffic features based on the image frame to obtain environmental traffic information;

[0064] S330: Inputting the environmental building information and the environmental traffic information into a pre-built inspection cycle analysis model to obtain a primary inspection cycle;

[0065] S340: Acquire target line location information and image acquisition time information based on the first environmental video information;

[0066] S350: Optimizing the primary inspection cycle based on the target line location information and the image acquisition time information to obtain a target line inspection cycle.

[0067] In one embodiment, the environmental building information and the environmental traffic information are input into a pre-built inspection cycle analysis model to obtain a primary inspection cycle, such as Figure 3 As shown, step S330 of the method provided by the present invention also includes:

[0068] S331: Acquire a training data set, where the training data set is multiple sets of environmental building data, environmental traffic data, and inspection cycle data;

[0069] S332: constructing the inspection cycle analysis model according to the BP neural network model, wherein the input data of the inspection cycle analysis model are the environmental building data and the environmental traffic data, and the output data is the inspection cycle;

[0070] S333: training, verifying and testing the inspection cycle analysis model based on the training data set, and generating the inspection cycle analysis model when the output accuracy of the inspection cycle analysis model meets the preset accuracy requirement;

[0071] S334: Input the environmental building information and the environmental traffic information into the inspection cycle analysis model to obtain the primary inspection cycle.

[0072] Specifically, in order to improve the scientific nature of the target line operation status inspection cycle, thereby effectively preventing obstacles from approaching the target line in power transmission operation and maintaining power transmission stability, this embodiment constructs the inspection cycle analysis model based on the BP neural network.

[0073] The input data of the inspection cycle analysis model are the environmental building data and the environmental traffic data, and the output data is the inspection cycle. The environmental building data is the building complexity feature data obtained based on the image building feature extraction, and the environmental traffic data is the traffic feature data obtained based on the image environment feature extraction, which is the traffic complexity feature of the target line and the transmission towers and poles. It should be understood that the higher the building complexity and the higher the environmental traffic complexity, the higher the possibility of the transmission line constructed in the scene being affected by the intrusion and causing operation failure, so the inspection cycle needs to be set shorter to ensure timely handling when the intrusion is entangled in the transmission line.

[0074] Based on historical inspection data, multiple groups of inspection data of power transmission lines built in different building environments and traffic environments are obtained to construct a training data set. The training data set is divided into a training set, a verification set and a test set according to a ratio of 7:2:1, and data identification is performed to facilitate data recognition by an inspection cycle analysis model. The inspection cycle analysis model is trained, verified and tested based on the training data set after identification. When the output accuracy of the inspection cycle analysis model meets the preset accuracy requirement, the inspection cycle analysis model is generated.

[0075] Architectural features are extracted based on the image frames to obtain the environmental architectural information; traffic features are extracted based on the image frames to obtain environmental traffic information; the environmental architectural information and the environmental traffic information are input into an inspection cycle analysis model; and the primary inspection cycle is output based on the inspection cycle analysis model. The primary inspection cycle is a cycle for which a drone is required to perform safety inspections on the area where the target line is located under the current environmental architectural conditions and surrounding environmental conditions.

[0076] It should be understood that in spring and summer when vegetation grows faster and in autumn when wind intensity is higher, target lines and towers and pole structures are prone to entanglement by plant growth and invasive objects (such as kites and greenhouse films), causing damage to transmission lines.

[0077] Therefore, in this embodiment, the geographical location of the target line is obtained based on the first environmental video information, that is, the target line location information is obtained, and the image acquisition time information of the current target line image acquisition is obtained based on the first environmental video information.

[0078] Based on the target line location information and the image acquisition time information, the latitude and longitude data of the area where the target line is located is determined, and combined with the image acquisition time to obtain the image acquisition season and the environmental wind conditions in the area, the primary inspection cycle is shortened and optimized (for example, the primary inspection cycle is shortened to 7 / 8 of the original cycle), and the target line inspection cycle is obtained. The target line inspection cycle is obtained based on the comprehensive analysis of the environmental building conditions, environmental traffic conditions, and dynamic target line construction environment affected by external intrusions in the static target line installation area. The dynamic change data is used to perform the next safety inspection on the target line. The target line inspection cycle can be dynamically adjusted according to the target line image acquisition time to meet the target line safety inspection guarantee needs under different seasonal conditions.

[0079] This embodiment collects historical data to construct an inspection cycle analysis model, and after obtaining the inspection cycle based on the inspection cycle analysis model, the inspection cycle is adjusted in combination with the current season and longitude and latitude information of the target line installation area, thereby achieving an inspection cycle that meets the target line safety inspection and protection needs under different conditions, thereby achieving the technical effect of performing scientific and reliable drone patrol and distance measurement on the target line.

[0080] In step S400, it is determined whether the image frame is an initial frame.

[0081] Specifically, it should be understood that target tracking is usually performed based on a single image. The present embodiment adopts a binocular image acquisition device to realize binocular image acquisition, and the first environmental video information and the second environmental video information are both characteristics of synchronous image acquisition of the same frame number for each target object in the same scene. Therefore, the present embodiment performs target tracking based on the image acquired by a single camera, and the target tracking object is the target line.

[0082] In the motion image data, each frame represents a still image of clothing, and the image information in adjacent continuous frames is similar. Therefore, this embodiment uses the initial frame as a key frame for performing target line tracking, and divides the target line into a region of interest (ROI) in the initial frame image to reduce the amount of data compared with the second video information during subsequent ranging.

[0083] In this embodiment, an image frame is acquired based on the first environmental video information, and whether the image frame is an initial frame is determined based on whether there are previous and subsequent frames in the image frame. If the image frame does not have a previous frame, the image frame is the initial frame, and the target line is positioned in two-dimensional space in the image frame based on the characteristics of the target transmission line.

[0084] In step S500, if the image frame is an initial frame, the target line is located in the image frame based on the characteristics of the target power transmission line to obtain target line location information.

[0085] like Figure 4 As shown, step S500 of the method provided by the present invention also includes:

[0086] S510: If the image frame is not an initial frame, randomly specifying a ROI candidate region in the image frame;

[0087] S520: extracting image features from the ROI candidate region to obtain candidate region features;

[0088] S530: performing a similarity comparison based on the candidate area features and the target transmission line features;

[0089] S540: Obtaining image similarity features;

[0090] S550: comparing the image similarity feature with a preset similarity threshold;

[0091] S560: If the image similarity feature meets the similarity threshold, a matching instruction is generated;

[0092] S570: Based on the matching instruction, call the ROI candidate area to perform area positioning matching in the second environment video information.

[0093] Specifically, in this embodiment, if the image frame is an initial frame, image feature comparison is performed between the image frame and the target power transmission line feature to obtain the location of the target line in the image frame and obtain the target line location information.

[0094] It should be understood that the distance measurement based on the drone in this embodiment is to measure the distance between the objects around the target line and the target line, and the purpose of the distance measurement is to ensure that there are no intrusions around the target line that interfere with the power transmission operation of the line. Therefore, this embodiment delineates the transmission line distance measurement area based on the drone image based on the target line positioning information, so as to reduce the image information comparison range when the second video information is subsequently combined for stereo matching and reduce the amount of calculation during stereo matching processing.

[0095] It should be understood that the initial frame is also a reference for subsequent frames to track the target line. Therefore, if the image frame is not the initial frame, it is necessary to locate the target line in the image frame to replace the initial frame with the key frame as the reference for subsequent frames to track the target line.

[0096] Therefore, if the image frame is not an initial frame, a certain image area in the image frame is randomly designated as a ROI candidate area; the same feature extraction method in step S100 is executed on the ROI candidate area to extract image features (extract image color features and line appearance features), obtain candidate area features, and perform similarity comparison between the candidate area features and the target power transmission line features, specifically, compare the similarity between the color features of the candidate area and the target power transmission line features and the similarity between the line appearance features of the candidate area and the target power transmission line features, and add the two similarities to obtain the image similarity feature. The larger the image similarity feature value, the more complete the ROI candidate area is for delineating the target line. Based on the ROI candidate area, stereo matching can be subsequently performed to measure the distance information between the intrusion around the target line and the target line.

[0097] The image similarity feature is compared with a preset similarity threshold. If the image similarity feature meets the similarity threshold, it indicates that the ROI candidate area is more complete in delineating the target line, and stereo matching can be performed based on the ROI candidate area. Therefore, a matching instruction is generated. Based on the matching instruction, the ROI candidate area is called to extract the corresponding image frame in the second environment video information, and regional positioning matching is performed on the extracted image. If the image similarity feature does not meet the similarity threshold, steps S510 to S570 are repeated until the image similarity feature meets the similarity threshold.

[0098] This embodiment randomly selects a ROI candidate area in an image frame to extract features and compare them with the features of the target power transmission line when the image frame is not an initial frame, so as to obtain an ROI area that satisfies the requirements for accurately and completely delineating the target line, thereby achieving the technical effect of reducing the amount of calculation for subsequent stereo matching and providing a reference benchmark for the tracking target of the next frame.

[0099] In step S600, a ROI region is extracted based on the target line location information;

[0100] In step S700, performing regional positioning matching on the second environment video information based on the ROI region to obtain ROI region matching information;

[0101] Specifically, in this embodiment, a region of interest (ROI region) is circled based on the target line positioning information, and subsequent two-dimensional image and three-dimensional imaging is performed on the ROI region. Based on the ROI region, image frames with the same number of frames are extracted from the second environmental video information for regional positioning matching to obtain ROI region matching information. The ROI region matching information and the ROI region are binocularly captured local images of the same target line and its environmental area.

[0102] In step S800, a stereo matching analysis is performed according to the target line positioning information, the ROI area and the ROI area matching information to obtain a target distance set.

[0103] See also Figure 5 , the method step S800 provided by the present invention further includes:

[0104] S810: Extract pixel point coordinates based on the ROI area to obtain a pixel point coordinate set;

[0105] S820: Extract pixel point coordinates based on the ROI area matching information to obtain a matching pixel point coordinate set;

[0106] S830: Inputting the pixel point coordinate set and the matching pixel point coordinate set into a pre-built stereo matching calculation model to obtain a three-dimensional scene in the ROI area;

[0107] S840: Inputting the target line positioning information into the ROI area three-dimensional scene to obtain the transmission line coordinate information;

[0108] S850: Obtaining environmental obstacle coordinate information based on the ROI area three-dimensional scene;

[0109] S860: Obtain the target distance set according to the environmental obstacle coordinate information and the transmission line coordinate information.

[0110] Specifically, in this embodiment, the three-dimensional coordinates of the object in the ROI area are restored based on binocular stereo vision technology, so as to measure the distance between the intruding object in the ROI area and the target line. Specifically, the existing two-dimensional image coordinate construction and spatial point depth information calculation method are used to extract the pixel point coordinates of the ROI area to obtain a pixel point coordinate set, and the pixel point coordinates are extracted based on the ROI area matching information to obtain a matching pixel point coordinate set.

[0111] The pixel point coordinate set and the matching pixel point coordinate set are input into a pre-built stereo matching calculation model to obtain a ROI area three-dimensional scene, wherein the ROI area three-dimensional scene is a three-dimensional modeling based on a local image of the ROI area, and the ROI area three-dimensional coordinate set of all pixels in the area can be obtained based on the ROI area three-dimensional scene. This implementation will be described in the subsequent description of the optimal embodiment of the stereo matching calculation model construction.

[0112] The target line positioning information is input into the ROI area three-dimensional scene to obtain a set of coordinate information of the target line in three-dimensional space, namely, the transmission line coordinate information. Based on the ROI area three-dimensional scene, the coordinate information of the environmental obstacles such as tree branches and kites that are not transmission lines in the ROI area is obtained.

[0113] Based on the environmental obstacle coordinate information and the transmission line coordinate information, the distance vectors between each environmental obstacle that interferes with the normal power transmission operation of the target line and the target line are calculated as the target distance set. The target distance set reflects the spatial distance between the environmental obstacle and the target line in the local three-dimensional coordinates constructed based on the ROI area, so as to achieve the technical effect of accurately knowing whether there are environmental obstacles that interfere with the normal power transmission operation of the line in the operating environment of the target line, thereby assisting operation and maintenance management personnel to perform transmission line maintenance management.

[0114] In one embodiment, the pixel point coordinate set and the matching pixel point coordinate set are input into a pre-built stereo matching calculation model to obtain a three-dimensional scene in the ROI area. The method step S830 provided by the present invention also includes:

[0115] S831: Constructing a stereo matching calculation model;

[0116] S832: The stereo matching calculation model includes a calibration parameter acquisition module, a calculation module and a three-dimensional scene generation module;

[0117] S833: The calculation module includes a disparity calculation unit and a depth distance calculation unit;

[0118] S834: Acquire camera calibration parameters of the binocular image acquisition device based on the calibration parameter acquisition module;

[0119] S835: Inputting the pixel point coordinate set and the matching pixel point coordinate set into the disparity calculation unit to obtain pixel point disparity;

[0120] S836: Inputting the pixel point disparity into the depth distance calculation unit to obtain a pixel point depth distance set;

[0121] S837: performing data processing on the pixel point depth distance set in the three-dimensional scene generation module to obtain a pixel point coordinate set;

[0122] S838: Generate the ROI area three-dimensional scene based on the pixel point coordinate set.

[0123] Specifically, in this embodiment, the preferred method for constructing the stereo matching calculation model is to construct a stereo matching calculation model including a calibration parameter acquisition module, a calculation module and a three-dimensional scene generation module, wherein the calculation module includes a disparity calculation unit and a depth distance calculation unit.

[0124] The calibration parameter acquisition module is used to acquire the camera calibration parameters of the binocular image acquisition device that performs the image acquisition task, including but not limited to focal length parameters and baseline length parameters, and the camera calibration parameters of the binocular image acquisition device are acquired based on the calibration parameter acquisition module.

[0125] The disparity calculation unit is used to calculate the disparity between the first video information image frame and the second video information image frame, and the pixel point coordinate set and the matching pixel point coordinate set are input into the disparity calculation unit for numerical calculation to obtain the pixel point disparity.

[0126] The pixel point parallax is input into the depth distance calculation unit, and the depth distance calculation unit calculates and obtains the pixel point depth distance set based on the principle of similar triangles. The pixel point depth distance set is processed in the three-dimensional scene generation module to obtain the spatial coordinate data of each pixel point in the world coordinate system to obtain the pixel point coordinate set, and the ROI area three-dimensional scene is generated based on the pixel point coordinate set, so as to provide a spatial distance calculation benchmark for the subsequent calculation of the spatial distance between the environmental obstacle and the target line based on the pixel point coordinate data, and to determine whether there is an environmental obstacle that causes the target line to have an operation risk.

[0127] Further, the target distance set is obtained according to the environmental obstacle coordinate information and the transmission line coordinate information, referring to Figure 7 , the method step S860 provided by the present invention further includes:

[0128] S861: preset distance threshold;

[0129] S862: Generate virtual distance coordinates according to the distance threshold and the transmission line coordinate information;

[0130] S863: traversing the environmental obstacle coordinate information based on the virtual distance coordinate to obtain a dangerous obstacle coordinate data set;

[0131] S864: Obtain the target distance set based on the dangerous obstacle coordinate data set and the transmission line coordinate information;

[0132] S865: performing risk level classification based on the target distance set and generating a risk level identification result;

[0133] S866: Perform a power transmission line safety warning based on the hazard level identification result.

[0134] Specifically, in this embodiment, the distance threshold is a safe distance at which environmental objects do not interfere with the normal operation of the transmission line, which can be obtained based on the experience of power department staff or based on safety codes. For example, when the transmission line is 10KV, the distance threshold is set to 3m.

[0135] Three-dimensional coordinate data is generated based on the distance threshold, and the transmission line coordinate information is added to generate virtual distance coordinates. The virtual distance coordinates can be regarded as a spatial area with the transmission line coordinates as the center of the circle. The distance between the edge coordinates of the area and any coordinate point in the transmission line coordinate information is greater than or equal to the distance threshold. If the environmental obstacle coordinate information of any environmental obstacle falls into the virtual distance coordinates, there is a risk of affecting the safe operation of the transmission line.

[0136] The environmental obstacle coordinate information is traversed based on the virtual distance coordinates to obtain the dangerous obstacle coordinate data set that falls within the virtual distance coordinates in the environmental obstacle coordinate information. Vector calculation is performed based on the dangerous obstacle coordinate data set and the transmission line coordinate information to obtain the target distance set of the straight-line distance between the dangerous obstacle coordinate data set and the transmission line. In the target distance set, the smaller the target distance data is, the smaller the spatial distance between the environmental obstacle and the target line is, and the greater the probability of entanglement and friction with the target line in the later stage, resulting in a power transmission safety accident of the target line.

[0137] A plurality of hazard level classification thresholds are preset, and based on the plurality of hazard level classification thresholds, the target distance set is traversed to classify the hazard level of each hazardous obstacle coordinate, and the hazardous obstacle coordinates are marked with a hazard level to obtain a hazard level marking result, and a transmission line safety warning is performed in three-dimensional imaging based on the hazard level marking result. The warning colors of different levels of hazards are different or have progressive colors, and color marking is performed in the three-dimensional scene of the ROI area.

[0138] This embodiment uses a safety distance threshold to add the coordinates of the transmission line to obtain the safety space that the transmission line should have with the environmental obstacles when it is in safe operation, thereby determining the coordinates of the environmental danger obstacles that pose a risk of interference to the operation of the target line and marking them in the three-dimensional scene of the ROI area. This achieves a technical effect of prompting power department workers to clear and remove environmental danger obstacles in the area where the target line is located step by step according to the danger level based on the image, thereby improving the timeliness of eliminating environmental dangers in the target line.

[0139] Figure 8 A schematic block diagram of the structure of a transmission line distance measurement system based on drone images in an embodiment of the present disclosure is shown.

[0140] like Figure 8 As shown, a transmission line distance measurement system based on drone images is provided, including: a feature extraction execution module 1, an image acquisition execution module 2, an image frame acquisition module 3, an image frame judgment module 4, a line feature positioning module 5, an interest area extraction module 6, an area matching execution module 7, and a target distance calculation module 8, wherein:

[0141] The feature extraction execution module 1 is used to extract the features of the target line and obtain the features of the target transmission line;

[0142] An image acquisition execution module 2 is used to acquire an image of a target area based on a binocular image acquisition device to obtain an environment video of the target area, wherein the environment video of the target area includes first environment video information and second environment video information;

[0143] An image frame acquisition module 3, used to acquire an image frame based on the first environment video information;

[0144] An image frame determination module 4 is used to determine whether the image frame is an initial frame;

[0145] A line feature positioning module 5, configured to locate the target line in the image frame based on the target power transmission line feature if the image frame is an initial frame, and obtain target line positioning information;

[0146] A region of interest extraction module 6 is used to extract a ROI region based on the target line positioning information;

[0147] A region matching execution module 7 is used to perform region positioning matching based on the ROI region in the second environment video information to obtain ROI region matching information;

[0148] The target distance calculation module 8 is used to perform stereo matching analysis according to the target line positioning information, the ROI area and the ROI area matching information to obtain a target distance set.

[0149] In one embodiment, the image frame obtaining module 3 further includes:

[0150] A building feature extraction unit, used to extract building features based on the image frame to obtain environmental building information;

[0151] A traffic feature extraction unit, used to extract traffic features based on the image frame to obtain environmental traffic information;

[0152] An inspection cycle obtaining unit, used for inputting the environmental building information and the environmental traffic information into a pre-built inspection cycle analysis model to obtain a primary inspection cycle;

[0153] An optimization information acquisition unit is used to acquire target line location information and image acquisition time information based on the first environmental video information;

[0154] The inspection cycle optimization unit is used to optimize the primary inspection cycle based on the target line position information and the image acquisition time information to obtain the target line inspection cycle.

[0155] In one embodiment, the inspection cycle obtaining unit further includes:

[0156] A training data acquisition unit, used to acquire a training data set, wherein the training data set is a plurality of sets of environmental building data, environmental traffic data, and inspection cycle data;

[0157] A model building execution unit, used to build the inspection cycle analysis model according to the BP neural network model, the input data of the inspection cycle analysis model is the environmental building data and the environmental traffic data, and the output data is the inspection cycle;

[0158] A model training execution unit, used for training, verifying and testing the inspection cycle analysis model based on the training data set, and generating the inspection cycle analysis model when the output accuracy of the inspection cycle analysis model meets the preset accuracy requirement;

[0159] The inspection cycle generating unit is used to input the environmental building information and the environmental traffic information into the inspection cycle analysis model to obtain the primary inspection cycle.

[0160] In one embodiment, the line feature positioning module 5 further includes:

[0161] A candidate region execution unit, configured to randomly specify a ROI candidate region in the image frame if the image frame is not an initial frame;

[0162] A region feature extraction unit, used to extract image features from the ROI candidate region to obtain candidate region features;

[0163] A feature similarity comparison unit, used for performing a similarity comparison based on the candidate area feature and the target transmission line feature;

[0164] A similarity value obtaining unit, used to obtain image similarity features;

[0165] A similarity threshold comparison unit, used to compare the image similarity feature with a preset similarity threshold;

[0166] A matching instruction generating unit, configured to generate a matching instruction if the image similarity feature satisfies a similarity threshold;

[0167] A matching instruction execution unit is used to call the ROI candidate area to perform area positioning matching in the second environment video information based on the matching instruction.

[0168] In one embodiment, the target distance calculation module 8 further includes:

[0169] A pixel coordinate obtaining unit, used for extracting pixel point coordinates based on the ROI area to obtain a pixel point coordinate set;

[0170] A pixel coordinate extraction unit, used to extract pixel point coordinates based on the ROI area matching information to obtain a matching pixel point coordinate set;

[0171] A three-dimensional scene generation unit, used for inputting the pixel point coordinate set and the matching pixel point coordinate set into a pre-built stereo matching calculation model to obtain a three-dimensional scene in the ROI area;

[0172] A line coordinate obtaining unit, used for inputting the target line positioning information into the ROI area three-dimensional scene to obtain the transmission line coordinate information;

[0173] An environmental coordinate obtaining unit, used to obtain environmental obstacle coordinate information based on the three-dimensional scene in the ROI area;

[0174] A target distance obtaining unit is used to obtain the target distance set according to the environmental obstacle coordinate information and the transmission line coordinate information.

[0175] Furthermore, the target distance obtaining unit further includes:

[0176] A distance threshold setting unit, used for presetting a distance threshold;

[0177] A virtual coordinate generating unit, configured to generate virtual distance coordinates according to the distance threshold and the transmission line coordinate information;

[0178] An obstacle coordinate obtaining unit, configured to traverse the environmental obstacle coordinate information based on the virtual distance coordinates to obtain a dangerous obstacle coordinate data set;

[0179] A target distance acquisition unit, configured to obtain the target distance set based on the dangerous obstacle coordinate data set and the transmission line coordinate information;

[0180] A danger level identification unit, used to classify danger levels based on the target distance set and generate a danger level identification result;

[0181] A safety warning execution unit is used to issue a safety warning for the power transmission line based on the hazard level identification result.

[0182] In one embodiment, the three-dimensional scene generation unit further includes:

[0183] A matching model generation unit, used for constructing a stereo matching calculation model;

[0184] The model module constituent unit, used for the stereo matching calculation model, includes a calibration parameter acquisition module, a calculation module and a three-dimensional scene generation module;

[0185] A calculation unit setting unit, used for the calculation module including a parallax calculation unit and a depth distance calculation unit;

[0186] A calibration parameter acquisition unit, used to acquire the camera calibration parameters of the binocular image acquisition device based on the calibration parameter acquisition module;

[0187] a disparity data calculation unit, configured to input the pixel point coordinate set and the matching pixel point coordinate set into the disparity calculation unit to obtain pixel point disparity;

[0188] A depth distance calculation unit, configured to input the pixel point disparity into the depth distance calculation unit to obtain a pixel point depth distance set;

[0189] A pixel coordinate generation unit, configured to perform data processing on the pixel point depth distance set in the three-dimensional scene generation module to obtain a pixel point coordinate set;

[0190] The three-dimensional scene construction unit is used to generate the three-dimensional scene of the ROI area based on the pixel point coordinate set.

[0191] For a specific embodiment of a transmission line ranging system based on drone images, please refer to the embodiment of a transmission line ranging method based on drone images mentioned above, which will not be repeated here. Each module in the above-mentioned transmission line ranging device based on drone images can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0192] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Fig. 9 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store news data and data such as time attenuation factors. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a transmission line distance measurement method based on drone images is implemented.

[0193] Those skilled in the art will understand that Fig. 9 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0194] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the following steps when executing the computer program: extracting features of a target line to obtain features of a target power transmission line; acquiring images of a target area based on the binocular image acquisition device to obtain an environmental video of the target area, wherein the environmental video of the target area includes first environmental video information and second environmental video information; acquiring an image frame based on the first environmental video information; determining whether the image frame is an initial frame; if the image frame is an initial frame, locating the target line in the image frame based on the features of the target power transmission line to obtain target line positioning information; extracting an ROI area based on the target line positioning information; performing regional positioning matching based on the ROI area in the second environmental video information to obtain ROI area matching information; performing stereo matching analysis based on the target line positioning information, the ROI area, and the ROI area matching information to obtain a target distance set.

[0195] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned transmission line distance measurement method based on drone images is implemented.

[0196] The parts not involved in the present invention are the same as the prior art or are implemented by using the prior art.

[0197] The above contents are further detailed descriptions of the present invention in combination with specific implementation methods, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the protection scope of the present invention.

Claims

1. A transmission line distance measurement method based on drone images, characterized in that: include: Extract features of the target line to obtain the features of the target transmission line; Capturing the target area image based on the binocular image acquisition device to obtain the target area environment video, wherein the target area environment video includes the first environment video information and the second environment video information; Acquire an image frame based on the first environmental video information; Determining whether the image frame is an initial frame; If the image frame is an initial frame, locating the target line in the image frame based on the target power transmission line feature to obtain target line location information; Extracting the ROI area based on the target line positioning information; Performing regional positioning matching based on the ROI region in the second environment video information to obtain ROI region matching information; Perform stereo matching analysis according to the target line positioning information, the ROI area and the ROI area matching information to obtain a target distance set; Wherein, acquiring an image frame based on the first environmental video information includes: Extracting building features based on the image frame to obtain environmental building information; Extracting traffic features based on the image frame to obtain environmental traffic information; Inputting the environmental building information and the environmental traffic information into a pre-built inspection cycle analysis model to obtain a primary inspection cycle; Acquire target line location information and image acquisition time information based on the first environmental video information; The primary inspection cycle is optimized based on the target line position information and the image acquisition time information to obtain a target line inspection cycle.

2. The transmission line distance measurement method based on drone images according to claim 1 is characterized in that: Inputting the environmental building information and the environmental traffic information into a pre-built inspection cycle analysis model to obtain a primary inspection cycle, including: Collect and obtain a training data set, wherein the training data set is multiple sets of environmental building data, environmental traffic data, and inspection cycle data; Constructing the inspection cycle analysis model according to the BP neural network model, wherein the input data of the inspection cycle analysis model are the environmental building data and the environmental traffic data, and the output data is the inspection cycle; Training, verifying and testing the inspection cycle analysis model based on the training data set, and generating the inspection cycle analysis model when the output accuracy of the inspection cycle analysis model meets the preset accuracy requirement; The environmental building information and the environmental traffic information are input into the inspection cycle analysis model to obtain the primary inspection cycle.

3. The transmission line distance measurement method based on drone images according to claim 1 is characterized in that: If the image frame is an initial frame, locating the target line in the image frame based on the target transmission line feature to obtain target line location information further includes: If the image frame is not an initial frame, randomly specifying a ROI candidate region in the image frame; Extracting image features from the ROI candidate region to obtain candidate region features; Performing a similarity comparison based on the candidate area features and the target transmission line features; Obtain image similarity features; Comparing the image similarity feature with a preset similarity threshold; If the image similarity feature meets the similarity threshold, a matching instruction is generated; Based on the matching instruction, the ROI candidate area is called to perform area positioning matching in the second environment video information.

4. The transmission line distance measurement method based on drone images according to claim 1 is characterized in that: Performing stereo matching analysis according to the target line positioning information, the ROI area and the ROI area matching information to obtain a target distance set includes: Extract pixel point coordinates based on the ROI area to obtain a pixel point coordinate set; Extract pixel point coordinates based on the ROI area matching information to obtain a set of matching pixel point coordinates; Input the pixel point coordinate set and the matching pixel point coordinate set into a pre-built stereo matching calculation model to obtain a three-dimensional scene in the ROI area; Input the target line positioning information into the ROI area three-dimensional scene to obtain the transmission line coordinate information; Obtaining environmental obstacle coordinate information based on the three-dimensional scene of the ROI area; The target distance set is obtained according to the environmental obstacle coordinate information and the transmission line coordinate information.

5. The transmission line distance measurement method based on drone images according to claim 4 is characterized in that: Inputting the pixel point coordinate set and the matching pixel point coordinate set into a pre-built stereo matching calculation model to obtain a three-dimensional scene in the ROI area, including: Construct a stereo matching computational model; The stereo matching calculation model includes a calibration parameter acquisition module, a calculation module and a three-dimensional scene generation module; The calculation module includes a disparity calculation unit and a depth distance calculation unit; Acquire the camera calibration parameters of the binocular image acquisition device based on the calibration parameter acquisition module; Inputting the pixel point coordinate set and the matching pixel point coordinate set into the disparity calculation unit to obtain pixel point disparity; Inputting the pixel point disparity into the depth distance calculation unit to obtain a pixel point depth distance set; In the three-dimensional scene generation module, data processing is performed on the pixel point depth distance set to obtain a pixel point coordinate set; The ROI region three-dimensional scene is generated based on the pixel point coordinate set.

6. The transmission line distance measurement method based on drone images according to claim 4 is characterized in that: Obtaining the target distance set according to the environmental obstacle coordinate information and the transmission line coordinate information includes: Preset distance threshold; Generate virtual distance coordinates according to the distance threshold and the transmission line coordinate information; Traversing the environmental obstacle coordinate information based on the virtual distance coordinates to obtain a dangerous obstacle coordinate data set; Obtaining the target distance set based on the dangerous obstacle coordinate data set and the transmission line coordinate information; Based on the target distance set, the danger level is divided and a danger level identification result is generated; A transmission line safety warning is performed based on the hazard level identification result.

7. A transmission line distance measurement system based on drone images, comprising: Feature extraction execution module, image acquisition execution module, image frame acquisition module, image frame judgment module, line feature positioning module, interest area extraction module, area matching execution module, target distance calculation module, among which: A feature extraction execution module is used to extract features of a target line and obtain features of a target transmission line; An image acquisition execution module is used to acquire an image of a target area based on a binocular image acquisition device to obtain an environment video of the target area, wherein the environment video of the target area includes first environment video information and second environment video information; An image frame acquisition module, used to acquire an image frame based on the first environment video information; The image frame acquisition module also includes: A building feature extraction unit, used to extract building features based on the image frame to obtain environmental building information; A traffic feature extraction unit, used to extract traffic features based on the image frame to obtain environmental traffic information; An inspection cycle obtaining unit, used for inputting the environmental building information and the environmental traffic information into a pre-built inspection cycle analysis model to obtain a primary inspection cycle; An optimization information acquisition unit is used to acquire target line location information and image acquisition time information based on the first environmental video information; An inspection cycle optimization unit, configured to optimize the primary inspection cycle based on the target line position information and the image acquisition time information to obtain a target line inspection cycle; An image frame determination module, used to determine whether the image frame is an initial frame; A line feature positioning module, configured to locate the target line in the image frame based on the target power transmission line feature if the image frame is an initial frame, and obtain target line positioning information; A region of interest extraction module, used to extract a ROI region based on the target line positioning information; A region matching execution module, used for performing region positioning matching based on the ROI region in the second environment video information to obtain ROI region matching information; The target distance calculation module is used to perform stereo matching analysis according to the target line positioning information, the ROI area and the ROI area matching information to obtain a target distance set.

8. A computer device, characterized in that: include: processor; as well as A memory, configured to store executable instructions of the processor; Wherein, the processor is configured to execute the transmission line ranging method based on drone images as described in any one of claims 1 to 6 by executing the executable instructions.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the transmission line distance measurement method based on drone images described in any one of claims 1 to 6 is implemented.

Images