Method and device for three-dimensional reconstruction of power transmission lines using cross lasers

Abstract

The present application provides a method and device for three-dimensional reconstruction of a power transmission line using cross lasers, which relates to the technical field of laser mapping. The method includes: extracting visual feature points through a cross light spot formed by the linkage control of a monocular camera and a laser emission device; simultaneously recording the trajectory coordinates when the monocular camera takes pictures to obtain the external parameters and internal parameters of the monocular camera; configuring a laser detection path and laser detection parameters based on the layout information of the power transmission line; determining the positional relationship between the cross light spot and the power transmission line, and limiting the relative spatial position between power transmission channels by constraining the cross spatial phase of the cross-line laser; fusing the point cloud data and visual feature points, and combining the monocular camera parameters, laser detection path and detection parameters to perform three-dimensional reconstruction of the power transmission line under the constraint of the cross spatial phase. The present application solves the technical problem that the existing three-dimensional reconstruction model is prone to cumulative errors, and improves the accuracy and reliability of the three-dimensional reconstruction of the power transmission line.

Description

Technical Field

[0001] This application relates to the technical field of laser mapping, and particularly to a method and device for three-dimensional reconstruction of transmission lines using cross lasers. Background Art

[0002] The three-dimensional reconstruction technology of transmission lines is an important means for monitoring, maintaining, and evaluating the status of transmission lines in the power industry. By constructing an accurate three-dimensional model, potential safety hazards, environmental impacts, and the actual operating status of transmission lines can be effectively detected, ensuring the stable operation of the power transmission system. Currently, the commonly used methods for three-dimensional reconstruction of transmission lines mainly include lidar and traditional photogrammetry. Lidar generates high-precision point cloud data by emitting laser pulses and receiving reflected signals, while traditional photogrammetry relies on multi-view images for feature point extraction and spatial reconstruction.

[0003] However, the surrounding of transmission lines is often accompanied by complex background environments, such as trees, buildings, mountains, etc. These backgrounds are easily confused with the characteristics of the lines themselves, making automatic feature extraction difficult. In complex environments, the reflection path of the laser is easily blocked, scattered, or reflected, resulting in incomplete or inaccurate data. Traditional photogrammetry lacks depth information, and its accuracy depends on the shooting angle and lighting conditions. In high-altitude and long-distance transmission lines, it is difficult to extract feature points, and the model stability is insufficient. In addition, both lidar and traditional photogrammetry lack effective spatial constraint means when dealing with large-scale scenes, and it is impossible to avoid cumulative errors in large-scale lines, making it difficult to ensure the relative spatial positions of each part of the model are exactly the same, resulting in a decrease in reconstruction accuracy. Summary of the Invention

[0004] This application provides a method and device for three-dimensional reconstruction of transmission lines using cross lasers, which solves the technical problems that the commonly used three-dimensional reconstruction technology of transmission lines is limited by complex environments and lacks effective spatial constraint means, resulting in insufficient accuracy of the three-dimensional reconstruction model and easy generation of cumulative errors, and achieves the technical effect of significantly improving the accuracy and reliability of the three-dimensional reconstruction of transmission lines.

[0005] In view of the above problems, on the one hand, the present application provides a method for three-dimensional reconstruction of a transmission line using cross lasers. The method includes: forming a cross light spot on the transmission line and its surrounding environment through the linkage control of a monocular camera and a laser emission device, and extracting visual feature points; at the same time, recording the trajectory coordinates when the monocular camera takes pictures to obtain the external parameters and internal parameters of the monocular camera; configuring a laser detection path and laser detection parameters based on the layout information of the transmission line; determining the positional relationship between the cross light spot mapped to the transmission line, and limiting the relative spatial position between transmission channels by constraining the cross spatial phase of the cross-line laser; fusing the point cloud data detected by the laser emission device and the visual feature points associated with the monocular camera, and combining the external parameters and internal parameters of the monocular camera, the laser detection path and laser detection parameters, and performing three-dimensional reconstruction of the transmission line under the constraint of the cross spatial phase.

[0006] On the other hand, the present application also provides a device for three-dimensional reconstruction of a transmission line using cross lasers. The device includes: a visual feature point extraction module for forming a cross light spot on the transmission line and its surrounding environment through the linkage control of a monocular camera and a laser emission device, and extracting visual feature points; a monocular camera parameter acquisition module for recording the trajectory coordinates when the monocular camera takes pictures to obtain the external parameters and internal parameters of the monocular camera; a laser detection configuration module for configuring a laser detection path and laser detection parameters based on the layout information of the transmission line; a relative spatial position limiting module for determining the positional relationship between the cross light spot mapped to the transmission line, and limiting the relative spatial position between transmission channels by constraining the cross spatial phase of the cross-line laser; a three-dimensional reconstruction module for fusing the point cloud data detected by the laser emission device and the visual feature points associated with the monocular camera, and combining the external parameters and internal parameters of the monocular camera, the laser detection path and laser detection parameters, and performing three-dimensional reconstruction of the transmission line under the constraint of the cross spatial phase.

[0007] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0008] Through the linkage control of a monocular camera and a laser emission device, a cross-shaped light spot is formed on the transmission line and the surrounding environment, and visual feature points are extracted. This step improves the data synchronization through the linkage control, making the positions of the feature points in space more accurate, avoiding the positioning errors that may be caused by solely relying on image feature points or laser data, and contributing to improving the overall reconstruction effect. At the same time, record the trajectory coordinates when the monocular camera takes pictures to obtain the external parameters and internal parameters of the monocular camera. This step determines the angle and distance changes during the camera movement through the accurate camera trajectory coordinates, thereby accurately correcting the spatial position of the camera and effectively improving the stability of spatial positioning in the 3D reconstruction of the transmission line. Based on the layout information of the transmission line, configure the laser detection path and laser detection parameters. This step dynamically configures the laser detection path and parameters according to the specific layout information of the transmission line, can flexibly cope with different line environments, dynamically configuring the detection path avoids the appearance of blind spots, and ensures the accuracy and coverage rate of detection in different scenarios, improving the accuracy of laser data. Determine the positional relationship between the cross-shaped light spot mapped to the transmission line, and limit the relative spatial position between the transmission channels by constraining the cross-space phase of the cross-line laser. This step effectively limits the relative spatial position between the channels of the transmission line by using the spatial phase constraint, greatly reducing the relative positioning error between the transmission lines and improving the overall accuracy and stability of the 3D reconstruction model. Integrate the point cloud data detected by the laser emission device and the visual feature points associated with the monocular camera, and combine the external parameters and internal parameters of the monocular camera, the laser detection path and laser detection parameters, and perform 3D reconstruction of the transmission line under the constraint of the cross-space phase. This step combines the fast trajectory recording of the camera and the real-time performance of laser detection, reduces the computational complexity, reduces the processing time for large-scale data, and significantly improves the efficiency of data acquisition and processing.

[0009] In summary, through means such as the linkage control of the monocular camera and the laser emission device, dynamic path configuration, and cross-shaped light spot space phase constraint, this application realizes high-precision 3D reconstruction of the transmission line, greatly improves the efficiency of data acquisition and processing, ensures the geometric accuracy and detail richness of the reconstruction model, significantly improves the accuracy and reliability of the 3D reconstruction of the transmission line, and at the same time enhances the adaptability in complex environments, providing strong technical support for the maintenance and safety management of the transmission line.

[0010] The above description is only an overview of the technical solution of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of this application more obvious and understandable, the specific embodiments of this application are specifically given below. Brief Description of the Drawings

[0011] Figure 1A schematic flow chart of a method for three-dimensional reconstruction of a power transmission line using a cross laser provided in an embodiment of the present application;

[0012] Figure 2 A schematic diagram of the structure of an apparatus for three-dimensional reconstruction of power transmission lines using a cross laser provided in an embodiment of the present application.

[0013] Explanation of the reference numerals: visualization feature point extraction module 10 , monocular camera parameter acquisition module 20 , laser detection configuration module 30 , relative spatial position limitation module 40 , three-dimensional reconstruction module 50 . DETAILED DESCRIPTION

[0014] The embodiments of the present application provide a method and device for three-dimensional reconstruction of transmission lines using a cross laser, thereby solving the technical problems that the commonly used three-dimensional reconstruction technology of transmission lines is limited by complex environments and lacks effective spatial constraints, resulting in insufficient accuracy of the three-dimensional reconstruction model and easy accumulation of errors, thereby achieving the technical effect of significantly improving the accuracy and reliability of three-dimensional reconstruction of transmission lines.

[0015] Embodiment 1, as Figure 1 As shown, the embodiment of the present application provides a method for three-dimensional reconstruction of a power transmission line using a cross laser, the method comprising:

[0016] Through the linkage control of the monocular camera and the laser emitting equipment, the cross spot formed on the transmission line and the surrounding environment is used to extract the visual feature points.

[0017] Specifically, the monocular camera is linked with the laser emitting device. The monocular camera is responsible for capturing the two-dimensional image of the transmission line and its environment, while the laser emitting device projects the laser in the target area to form a cross spot, such as a lidar or laser scanner. Through the precise positioning of the cross laser, the key parts of the line, namely the feature points, can be clearly identified. These feature points are unique and easy to identify points in the image. Insulators on the transmission line, nodes of the tower, and wire suspension points can all be regarded as feature points.

[0018] In the 3D reconstruction of power transmission lines, the laser emission equipment forms a clear cross spot on the power transmission line and its surrounding environment. The image of these cross spots is captured by a monocular camera, and the visual feature points are extracted in combination with the position information of the cross spots. The extracted feature points are then used for further 3D reconstruction and analysis. The visual feature points extracted through this linkage control provide key data for subsequent spatial positioning and model generation, improving the accuracy of the entire reconstruction process.

[0019] At the same time, the trajectory coordinates of the monocular camera during shooting are recorded to obtain the external parameters and internal parameters of the monocular camera.

[0020] Specifically, the trajectory coordinates are a dataset of coordinates formed by a monocular camera continuously recording its position and angle changes during movement, reflecting the camera's movement trajectory. When photographing a transmission line, the monocular camera moves along a preset path. While moving, the camera continuously captures images of the surrounding environment and records every position and attitude change of the camera through a built-in positioning system, IMU (Inertial Measurement Unit), etc. These data constitute the trajectory coordinates.

[0021] Using the trajectory coordinates, the rotation and translation of the monocular camera relative to the world coordinate system at each moment are calculated, thereby obtaining the external parameters of the monocular camera. The external parameters refer to the position and orientation of the monocular camera relative to the world coordinate system, such as the rotation matrix and translation vector.

[0022] Before the monocular camera performs the shooting task, the monocular camera is calibrated to determine its internal parameters. The internal parameters include the focal length of the monocular camera, the position of the principal point, and the lens distortion parameters. The calibration process can be carried out by photographing a calibration plate (such as a checkerboard) with known size and features, and then using image processing algorithms to calculate the focal length, principal point position, and distortion parameters of the camera.

[0023] By recording the trajectory coordinates of the monocular camera and combining camera calibration techniques, the external and internal parameters of the camera can be accurately obtained, realizing an accurate description of the camera's position and characteristics, providing key data support for subsequent 3D model generation, and improving the accuracy and reliability of the overall reconstruction.

[0024] Based on the layout information of the transmission line, configure the laser detection path and laser detection parameters.

[0025] Specifically, the layout information of the transmission line refers to the detailed data of the position, orientation, structure, and related facilities (such as substations, supports, etc.) of the transmission line in the geographical space. These information usually include the height, angle, spacing of the line, and environmental factors. Collect the detailed layout information of the transmission line. This can be obtained through existing engineering drawings or geographical information system (GIS) data, etc. Through the layout information, the orientation, height, and surrounding environment characteristics of the line can be understood, and possible obstacles, signal interference sources, etc. can be analyzed. For example, if there are trees near the line, it is necessary to avoid the laser beam being blocked.

[0026] Based on the acquired layout information and environmental analysis, a laser detection path is formulated. The configuration of the laser detection path needs to consider factors such as the length of the transmission line, terrain undulation, vegetation coverage, etc. The path should follow the center line of the transmission line or be appropriately offset to ensure that the laser emission device can capture detailed information about the transmission towers, conductors, and the surrounding environment. For example, if the transmission line is in a large valley, the detection path may need to be set along both sides of the valley to obtain the best measurement results. After determining the detection path, laser detection parameters also need to be configured. According to the distance to the target, environmental conditions, and required measurement accuracy, adjust the laser's emission power, frequency, and angle. For example, in long-distance measurements, it may be necessary to increase the laser's emission power to ensure that the signal can reach the target and be reflected back to the receiver.

[0027] Through the above steps, it is possible to achieve a comprehensive measurement and analysis of the transmission line, improve the efficiency and accuracy of laser detection, and provide accurate laser measurement data for subsequent 3D reconstruction.

[0028] Determine the positional relationship between the cross-shaped light spot mapped to the transmission line, and limit the relative spatial position between transmission channels by constraining the cross-spatial phase of the cross-line laser.

[0029] Specifically, the cross-spatial phase refers to the specific configuration and arrangement of the cross-shaped light spot in three-dimensional space, including the relative position and angle between the light spots.

[0030] Process the images captured by the aforementioned monocular camera. Through image feature point detection and matching techniques, such as SIFT (Scale-Invariant Feature Transform), SURF (Speeded-Up Robust Features), etc., identify the position of the cross-shaped light spot in the image. Combining the internal and external parameters of the camera, convert the two-dimensional position of the light spot into the actual position in three-dimensional space, thereby determining the position of the cross-shaped light spot in the actual transmission line.

[0031] In order to limit the relative spatial position between transmission channels, it is necessary to constrain the phase of the cross-shaped light spot, set the shape and arrangement of the cross-shaped light spot, and ensure its consistency and stability in space. For example, it can be required that the intersection point of the cross-shaped light spot always be located at the midline position of the transmission line to ensure that there is no error caused by the offset of the light spot during measurement. By constraining the cross-spatial phase, the relative spatial position between transmission channels can be accurately limited. For example, at a certain intersection point, two transmission lines may intersect. By constraining the spatial phase of the cross-shaped light spot, to ensure the accurate positioning of the intersection point during measurement.

[0032] Through the above steps, during the 3D reconstruction process of the transmission line, it is possible to achieve precise positioning of key points and effective control of relative positions, further improving the accuracy and reliability of 3D reconstruction.

[0033] Fuse the point cloud data detected by the laser emission device and the visual feature points associated with the monocular camera, and combine the external parameters and internal parameters of the monocular camera, the laser detection path and the laser detection parameters. Under the constraint of the cross-space phase, perform 3D reconstruction of the transmission line.

[0034] Specifically, pair the point cloud data obtained by the laser emission device with the visual feature points associated with the monocular camera to ensure that the position of each feature point in the 3D space can correspond to the corresponding point cloud data. The point cloud data provides 3D space information, while the visual feature points serve as key reference points. The fusion of the two provides a comprehensive data basis for 3D reconstruction. Use the external and internal parameters of the monocular camera, combined with the laser detection parameters, to calibrate and register the fused data to ensure that all data is within a unified coordinate system. During the fusion process, the constraint of the cross-space phase is adopted to ensure that all point cloud data and feature points maintain a preset spatial relationship, avoiding misalignment caused by data processing. After completing data fusion and parameter combination, use 3D reconstruction algorithms such as Structure from Motion (SfM), Multi-View Geometry (MVG), or Bundle Adjustment to process the fused point cloud and feature point data to generate a 3D model of the transmission line to visually display the actual situation of the transmission line.

[0035] Further, in the embodiment of the present application, through the linkage control of the monocular camera and the laser emission device, cross-shaped light spots are formed on the transmission line and the surrounding environment, and visual feature points are extracted, including:

[0036] Filter the cross-shaped light spots formed on the transmission line and the surrounding environment to obtain cross-shaped light spot feature points; based on the cross-shaped light spot feature points, train a light spot classification model, and the light spot classification model is used to filter non-target light spots; use the light spot classification model to retain target light spots and extract the visual feature points.

[0037] Specifically, the laser emission device generates cross-shaped light spots in the transmission line and its surrounding environment, and the monocular camera captures images formed by these light spots in real time. Preprocess the captured images through operations such as grayscale conversion, noise reduction, and contrast enhancement to highlight the features of the cross-shaped light spots and identify and extract all cross-shaped light spots. Refine the filtered cross-shaped light spots to extract their precise position, shape, size and other features, and these feature points will be used as the basis for subsequent training and classification.

[0038] Based on the identified cross light spot feature points, a large amount of image data containing target light spots and non-target light spots is collected, annotated and classified as a training data set. Among them, the target light spot refers to the cross light spot located on the transmission line, and the non-target light spot is a light spot that is not related to the transmission line, such as background noise or reflection of the surrounding environment. Use deep learning frameworks such as TensorFlow or PyTorch to build and train a light spot classification model. The light spot classification model is used to identify and classify cross light spots and distinguish between target light spots and non-target light spots. Common model structures include convolutional neural networks (CNNs) and recurrent neural networks (RNNs). Use the labeled data set to train the model, adjust the network weights through the back propagation algorithm to minimize the difference between the predicted results and the actual labels, and use cross-validation technology to verify the performance of the model on the training set and validation set. According to the validation results, the model parameters are continuously adjusted until the model converges to obtain a trained light spot classification model.

[0039] After the training is completed, the spot classification model is applied to analyze the new spot data. The real-time or pre-processed image is input into the trained spot classification model, and the model will output the classification result of each spot. According to the results output by the model, the feature points identified as the target spot are retained, and these feature points will serve as the key information for subsequent 3D reconstruction.

[0040] The above steps can efficiently and accurately extract visual feature points in the transmission line and the surrounding environment by training the spot classification model, thereby improving the accuracy of the three-dimensional reconstruction data.

[0041] Furthermore, in the steps of the embodiment of the present application, the monocular camera and the laser emitting device are controlled in linkage to perform timing synchronization through a synchronous control gate circuit; based on the external parameters and internal parameters of the monocular camera, a first acquisition window is determined, and the first acquisition window includes a first rotation angle; based on the laser detection path and laser detection parameters of the laser emitting device, a first light spot coverage area is determined, and the first light spot coverage area includes a second rotation angle; according to the synchronous control gate circuit, in combination with the first acquisition window and the first light spot coverage area, the monocular camera and the laser emitting device are controlled in linkage.

[0042] Specifically, a synchronous control gate circuit is used to synchronize the operation of the monocular camera and the laser emitting device to ensure that the image acquisition of the monocular camera and the light spot emission of the laser emitting device are precisely aligned in time. The external parameters and internal parameters of the monocular camera are used to determine the field of view and direction of the monocular camera at a specific time point, which is defined as the first acquisition window, which contains the first rotation angle. The first rotation angle is the rotation angle of the monocular camera when acquiring images, which is used to determine the range of image acquisition. According to the laser detection path and laser detection parameters of the laser emitting device, the cross light spot coverage area formed by the laser on the transmission line and the environment is calculated, that is, the first light spot coverage area, which includes the second rotation angle. The second rotation angle is the rotation angle of the laser emitting device when emitting lasers, which is used to determine the range of light spot coverage. Combined with the synchronous control gate circuit, the first acquisition window and the first light spot coverage area, the control parameters of the laser emitting device and the monocular camera are adjusted to ensure that the laser emitting device generates a light spot in the corresponding area while the camera is acquiring images.

[0043] Through synchronous control, it can be ensured that the obtained images and laser data can be accurately combined to collect accurate and effective data for subsequent three-dimensional reconstruction.

[0044] Furthermore, the embodiment of the present application controls the monocular camera and the laser emission device in linkage according to the synchronous control gate circuit, in combination with the first acquisition window and the first light spot coverage area, including:

[0045] Monitor the changes in ambient light in real time and obtain external lighting conditions. At the same time, evaluate the clarity based on the point cloud data detected by the laser emitting device and the visual feature points associated with the monocular camera, where the clarity includes the density of the visual feature points and the density of the point cloud data. Based on the external lighting conditions, dynamically adjust the exposure time of the monocular camera and the laser intensity of the laser emitting device in combination with the clarity.

[0046] Specifically, an ambient light sensor is deployed around the monocular camera and laser emitting device, and the ambient light sensor is used to monitor the light intensity in real time to obtain external lighting conditions, including information such as light direction, intensity, and color temperature. By combining the point cloud data obtained by the laser emitting device and the visual feature points associated with the monocular camera, the image clarity is evaluated by calculating the density of the visual feature points and the density of the point cloud data, and a quantitative clarity index is obtained. Among them, the feature point density refers to the number of visual feature points per unit area, which is obtained by calculating the ratio of the number of feature points to the image area; the point cloud data density refers to the number of point cloud data points per unit volume, which is used to evaluate the fineness and coverage of the scan, and can be obtained by calculating the ratio of the number of point cloud points to the scan volume. For complex scenes, it needs to be estimated through a point cloud processing algorithm.

[0047] During the 3D reconstruction process, according to the real-time external lighting conditions and the evaluation results of image clarity, the parameters of the monocular camera and the laser device are adjusted through a PID (Proportional-Integral-Derivative) controller or fuzzy logic, and then the exposure time of the monocular camera and the laser intensity of the laser emission device are adjusted. In a low-light environment, the exposure time is increased to obtain more light signals; in a high-light environment, the exposure time is reduced to prevent overexposure. When the density of the point cloud data is insufficient, the laser intensity is increased to improve the scanning accuracy; when the density of the feature points is insufficient, the number of feature points is increased by adjusting the laser mode or frequency.

[0048] Through real-time monitoring and dynamic adjustment strategies, the monocular camera and the laser emission device can adapt to various environmental conditions. Even in an environment where the lighting conditions change rapidly, the clarity and quality of the images and point cloud data can be guaranteed, providing stable data support for subsequent data analysis and 3D reconstruction.

[0049] Furthermore, based on the external lighting conditions, the exposure time of the monocular camera and the laser intensity of the laser emission device are dynamically adjusted in combination with the clarity in the embodiments of the present application, and it further includes:

[0050] Through the density of the visual feature points in the clarity, a first sparsely distributed partition and a first densely distributed partition are obtained; through the density of the point cloud data in the clarity, a second sparsely distributed partition and a second densely distributed partition are obtained; based on the layout information of the transmission line, key line nodes are obtained, and the key line nodes are located by constraining the spatial complexity; based on the first sparsely distributed partition, the first densely distributed partition, the second sparsely distributed partition, and the second densely distributed partition, in comparison with the key line nodes, clarity constraint intervals are added.

[0051] Specifically, according to the density of the visual feature points in the clarity, regions with a lower density of feature points are identified and denoted as the first sparsely distributed partition, and regions with a higher density are denoted as the first densely distributed partition. For example, a threshold range for the density of the visual feature points is set. If the density of the feature points in a certain region is lower than the minimum value of the set threshold range, it is marked as a sparse partition, and if it is higher than the maximum value of the set threshold range, it is marked as a dense partition. By analyzing the density of the point cloud data in the clarity, regions where the point cloud data is sparsely distributed are determined and denoted as the second sparsely distributed partition, and regions where the point cloud data is densely distributed are denoted as the second densely distributed partition.

[0052] Combined with the layout information of the transmission line, through spatial complexity constraints, locate the key nodes of the line. These nodes are important position points in the transmission line, such as wire connection points, towers, etc. Among them, the spatial complexity constraint is a constraint condition considered when locating the key nodes of the line in three-dimensional space, usually related to the structure and layout of the line, such as line turning points, intersection points, etc. Compare the key nodes of the line with the feature points and the density partition of the point cloud data to identify which key nodes are located in the sparse area. Add a clarity constraint interval around these key nodes to ensure the data quality in these areas. By adjusting the exposure time of the monocular camera and the laser intensity of the laser emission device, enhance the feature points and the density of the point cloud data in this area. Increase the exposure time in the first sparse distribution partition where the key node is located to improve the recognition rate of the feature points. In the second sparse distribution partition where the key node is located, increase the laser intensity to improve the density of the point cloud data and ensure the integrity and accuracy of the data.

[0053] Through the above partition-based clarity constraint, the high-precision detection of the key area of the transmission line is ensured, and the clarity and quality of the key data are improved.

[0054] Furthermore, the method described in the embodiment of the present application further includes:

[0055] Based on the layout information of the transmission line, mark the distribution coordinates of each monocular camera to obtain the monocular camera distribution mark; according to the monocular camera distribution mark, compare with the laser detection path to determine the monocular camera collaborative activation sequence; based on the monocular camera collaborative activation sequence and the monocular camera distribution mark, match the monocular camera cross-activation sequence, and the monocular camera collaborative activation sequence corresponds one-to-one with the monocular camera cross-activation sequence.

[0056] Specifically, the monocular camera collaborative activation sequence is the camera activation order determined according to the distribution mark of the monocular camera and the laser detection path to ensure full coverage and efficient data acquisition. The monocular camera cross-activation sequence is another activation order that matches the collaborative activation sequence and is used to achieve cross-verification and data supplementation between different cameras.

[0057] Based on the layout information of the transmission line, including the line's orientation, height, surrounding environment, etc., determine the distribution positions and angles of each monocular camera. Mark the distribution coordinates of each monocular camera, record its position parameters, and form a monocular camera distribution mark, providing basic information for subsequent collaborative activation. Compare with the detection path of the laser emission device, analyze its intersection points and key areas with the transmission line layout information, and determine the collaborative activation order of the monocular cameras based on the intersection points and key areas, forming a monocular camera collaborative activation sequence to ensure that when the laser spot forms in the key area, the corresponding monocular camera can capture it in a timely manner. Based on the monocular camera collaborative activation sequence, design a cross-activation strategy to achieve more comprehensive coverage and more refined data collection through the cooperation between adjacent or diagonally distributed cameras. Match a corresponding cross-activation sequence for each collaborative activation sequence to form a one-to-one correspondence. When a group of cameras works according to the collaborative activation sequence, another group of cameras works according to the cross-activation sequence, thereby realizing cross-verification and supplementation of data.

[0058] Exemplarily, when monitoring a transmission line, assign coordinates to three monocular cameras through the layout information, which are A(0, 0, 0), B(5, 0, 0), and C(10, 0, 0) respectively. After obtaining these distribution marks, determine the activation sequence as "Camera A - Camera B - Camera C" according to the laser detection path. Next, match the cross-activation sequence based on this activation sequence. For example, the activation order is "Camera A is activated first, then Camera B is activated, and at the same time Camera A is activated for cross", and finally form a sequence of "Camera A - Camera B - Camera A - Camera C".

[0059] Through this coordinated activation method, it can ensure the effective utilization of light and laser detection data between different cameras, improve the coverage and quality of data, and provide support for subsequent 3D reconstruction.

[0060] Furthermore, the monocular camera collaborative activation sequence and the monocular camera cross-activation sequence in the embodiment of the present application are in one-to-one correspondence, and further include:

[0061] Based on the laser detection path, combine the second rotation angle of the first light spot coverage area to generate a scanning coverage space; in the scanning coverage space, combine the first rotation angle of the first acquisition window to locate the first collaborative monocular camera in the monocular camera collaborative activation sequence; use the first collaborative monocular camera as the center, and according to the monocular camera distribution mark, determine multiple first cross-monocular cameras, and the first cross-monocular cameras are located with time complexity constraints; according to the first collaborative monocular camera and multiple first cross-monocular cameras, traverse the monocular camera collaborative activation sequence to determine the monocular camera cross-activation sequence.

[0062] Specifically, for the time complexity constraint: The time factor considered when determining the first cross monocular camera ensures the efficiency and real-time nature of data acquisition.

[0063] Based on the laser detection path and combined with the second rotation angle of the first spot coverage area, a coverage space for laser scanning is generated, that is, the movement trajectory and range of the laser spot in space. In the scanned coverage space, using the first rotation angle of the first acquisition window, a monocular camera capable of effectively capturing the spot is located and determined as the first collaborative monocular camera. This first collaborative monocular camera is a monocular camera located according to the scanned coverage space and the first rotation angle of the first acquisition window and is responsible for leading the data acquisition monocular camera. Centered on the first collaborative monocular camera, according to the monocular camera distribution marker, through the time complexity constraint, that is, ensuring that within the laser spot coverage range, the camera can respond in a timely manner and capture the spot, multiple first cross monocular cameras are determined. Among them, the first cross monocular camera is another monocular camera that works in cooperation with the first collaborative monocular camera and is responsible for alternating work at different time points to enhance the data acquisition effect. Traverse the monocular camera collaborative activation sequence. For each first collaborative monocular camera, determine its corresponding multiple first cross monocular cameras to form a one-to-one corresponding monocular camera cross activation sequence.

[0064] Exemplarily, in the monitoring of a certain power transmission line, the detection path of the laser emission device determines a spot coverage area, and the second rotation angle is 30 degrees. Based on this information, a conical scanned coverage space is generated, which contains multiple data acquisition points. Within this scanned coverage space, the first rotation angle of the first acquisition window is 45 degrees. By analyzing this area, "Camera A" is determined as the first collaborative monocular camera, which is responsible for the main data acquisition task. Next, taking "Camera A" as the center point, according to the camera distribution marker and considering the time complexity at the same time to ensure that these cross cameras can be activated within the corresponding time window and complete data acquisition, "Camera B" and "Camera C" are determined as the first cross monocular cameras.

[0065] Through the above steps, it can be ensured that during the entire data acquisition process, each camera can cooperate effectively, maximize the coverage of the monitoring area, improve the data quality, and provide comprehensive and accurate data support for subsequent three-dimensional reconstruction.

[0066] Furthermore, the embodiment of the present application determines the positional relationship between the cross spot mapped to the power transmission line and, by constraining the cross space phase of the cross-line laser, limits the relative spatial position between power transmission channels, and further includes:

[0067] In the scanning coverage space, eight quadrant sub-spaces are set by constraining the cross-space phase of the crosshair laser; based on the eight quadrant sub-spaces and the first collaborative monocular camera, relative position encoding is performed on multiple first cross monocular cameras, wherein the first collaborative monocular camera is assigned an encoding value according to the positional relationship of the eight quadrant sub-spaces; based on the first collaborative monocular camera and multiple first cross monocular cameras, through the relative position encoding, a spatial index for the three-dimensional reconstruction of the transmission line is established.

[0068] Specifically, the laser emission device emits crosshair lasers, and these laser lines cross in space to form a cross-shaped light spot. This cross light spot is the basis for spatial division. By adjusting the laser emission device, the intersection point and direction of the laser lines are precisely controlled so that the distribution of the crosshair lasers in space meets specific phase requirements to ensure that the position and shape of the light spot in space conform to expectations. Based on the distribution of the crosshair lasers, the scanning coverage space is divided into eight quadrant sub-spaces. Each quadrant sub-space represents a part of the cross light spot, similar to each corner in an octagonal structure. Each sub-space has clear boundaries and coordinates, providing a spatial framework for subsequent position encoding and three-dimensional reconstruction.

[0069] Based on the positional relationship of the eight quadrant sub-spaces, the specific position of the first collaborative monocular camera in space is determined, including the relative coordinates of the camera in the quadrant, as well as the distance and angle between the camera and the light spot formed by the crosshair laser. A specific encoding value is assigned to the first collaborative monocular camera, and this encoding value contains the accurate position information of the camera in space. The encoding can be in the form of numbers, letters, or binary, depending on the design of the encoding system. Taking the first collaborative monocular camera as the center, a relative coordinate system is established. This coordinate system can be a Cartesian coordinate system or a polar coordinate system. Define the encoding rules to ensure that the encoding value of each first cross monocular camera can accurately reflect its relative position relationship with the first collaborative monocular camera. The encoding rules need to ensure the uniqueness and scalability of the encoding. For multiple first cross monocular cameras, through geometric calculations, their relative distances, angles, and spatial positions with respect to the first collaborative monocular camera are determined for relative position encoding. The encoding value of each cross monocular camera reflects its position relationship in the relative coordinate system.

[0070] Exemplarily, in one monitoring, the first collaborative monocular camera is placed in Quadrant 1 and is assigned a coding value of "1". This first collaborative monocular camera has two first cross monocular cameras, which are located in Quadrant 2 and Quadrant 3 respectively. The relative relationship between the first cross monocular camera A (located in Quadrant 2) and the first collaborative monocular camera is: distance: 2 meters, angle: 30 degrees. Therefore, the coding value assigned to this camera is "2-2-30". The relative relationship between the first cross monocular camera B (located in Quadrant 3) and the first collaborative monocular camera is: distance: 3 meters, angle: 45 degrees. Accordingly, the coding value of this camera is "3-3-45". Through such a coding system, the positional relationship of each camera in space can be clearly described, enabling quick positioning and acquisition of relevant data during subsequent data processing and 3D reconstruction processes.

[0071] According to the relative position coding determined above, collect the position information of all first cross monocular cameras relative to the first collaborative monocular camera. These codings reflect the specific positions and orientations of each camera in the eight quadrant subspaces. Create a data structure to store and manage these coding information, such as an array, a list, a hash table, or a tree structure. In the selected data structure, organize the data of the first cross monocular cameras according to the relative position coding. For example, the coding value can be used as an index or a key value, and the position and orientation information of the camera can be used as the corresponding value or node. Establish an index mapping to associate the relative position coding of each camera with its position in the dataset, and obtain the 3D reconstruction spatial index of the transmission line. This index corresponds the positions of each camera and the collected data to the actual structure of the transmission line, facilitating quick and accurate positioning and access to different spatial data during the reconstruction process.

[0072] Through the relative position coding, combined with the collaborative work of the first collaborative monocular camera and the first cross monocular cameras, not only the accuracy and efficiency of data collection are improved, but also key spatial information for the 3D reconstruction of the transmission line is provided to improve the accuracy and reliability of the 3D reconstruction model.

[0073] In summary, the method for 3D reconstruction of a transmission line using a cross laser provided by the embodiments of the present application has the following technical effects:

[0074] The monocular camera and the laser emitting device are controlled by the synchronous control gate circuit to form a cross light spot on the transmission line and the surrounding environment, train the light spot classification model, and extract the visual feature points. This step improves the data synchronization through linkage control, making the position of the feature points in space more accurate, avoiding the positioning error that may be caused by relying solely on image feature points or laser data, and helping to improve the overall reconstruction effect. In addition, combined with the external lighting conditions and clarity, the exposure time of the monocular camera and the laser intensity of the laser emitting device are adjusted in comparison with the key nodes of the transmission line to obtain accurate and reliable key point data and improve the accuracy of the unit reconstruction data. At the same time, the trajectory coordinates of the monocular camera when shooting are recorded to obtain the external parameters and internal parameters of the monocular camera. This step determines the angle and distance changes during the camera movement through the precise camera trajectory coordinates, thereby accurately correcting the spatial position of the camera, effectively improving the stability of spatial positioning in the three-dimensional reconstruction of the transmission line. Based on the layout information of the transmission line, the laser detection path and laser detection parameters are configured. This step dynamically configures the laser detection path and parameters according to the specific layout information of the transmission line, which can flexibly respond to different line environments. The dynamic configuration of the detection path avoids the appearance of blind spots, ensures the accuracy and coverage of detection in different scenarios, and improves the accuracy of laser data. Set the monocular camera collaborative activation sequence and the monocular camera cross activation sequence to cross-validate the collected data, further improve the accuracy of the unit reconstruction data, and reduce data errors. Determine the positional relationship between the cross spot mapped to the transmission line, and by constraining the cross spatial phase of the cross-line laser, limit the relative spatial position between the transmission channels, and establish a spatial index for the three-dimensional reconstruction of the transmission line. This step uses the spatial phase constraint to effectively limit the relative spatial position between the channels of the transmission line, greatly reducing the relative positioning error between the transmission lines, and improving the overall accuracy and stability of the three-dimensional reconstruction model. The point cloud data detected by the laser emitting device and the visual feature points associated with the monocular camera are fused, and the external and internal parameters of the monocular camera, the laser detection path and the laser detection parameters are combined to perform three-dimensional reconstruction of the transmission line under the constraint of the cross spatial phase. This step combines the camera's fast trajectory recording with the real-time laser detection, reduces the computational complexity, reduces the processing time for large-scale data, and significantly improves the efficiency of data acquisition and processing.

[0075] In general, the embodiments of the present application achieve high-precision three-dimensional reconstruction of transmission lines by means of linkage control of monocular cameras and laser emitting equipment, dynamic path configuration, and cross-spot spatial phase constraints, which greatly improves the efficiency of data acquisition and processing, ensures the geometric accuracy and detail richness of the reconstructed model, significantly improves the accuracy and reliability of three-dimensional reconstruction of transmission lines, and enhances the adaptability in complex environments, providing strong technical support for the maintenance and safety management of transmission lines.

[0076] Embodiment 2, as Figure 2 shown, the embodiment of the present application provides a device for three-dimensional reconstruction of a transmission line using a cross laser. The device includes:

[0077] A visual feature point extraction module 10, which is used to extract visual feature points by controlling the linkage of a monocular camera and a laser emission device to form a cross light spot on the transmission line and its surrounding environment.

[0078] A monocular camera parameter acquisition module 20, which is used to record the trajectory coordinates when the monocular camera takes pictures to obtain the external parameters and internal parameters of the monocular camera.

[0079] A laser detection configuration module 30, which is used to configure the laser detection path and laser detection parameters based on the layout information of the transmission line.

[0080] A relative spatial position limitation module 40, which is used to determine the positional relationship between the cross light spot mapped to the transmission line and limit the relative spatial position between transmission channels by constraining the cross spatial phase of the cross-line laser.

[0081] A three-dimensional reconstruction module 50, which is used to fuse the point cloud data detected by the laser emission device and the visual feature points associated with the monocular camera, and perform three-dimensional reconstruction of the transmission line under the constraint of the cross spatial phase in combination with the external parameters and internal parameters of the monocular camera, and the laser detection path and laser detection parameters.

[0082] Furthermore, the visual feature point extraction module 10 of the embodiment of the present application is further used to perform the following steps:

[0083] Filter the cross light spot formed on the transmission line and its surrounding environment to obtain cross light spot feature points; train a light spot classification model based on the cross light spot feature points, and the light spot classification model is used to filter non-target light spots; use the light spot classification model to retain the target light spots and extract the visual feature points.

[0084] Further, in the visual feature point extraction module 10 of the embodiment of the present application, the monocular camera and the laser emission device are linked and controlled to synchronously control the gate circuit for timing synchronization; based on the external parameters and internal parameters of the monocular camera, a first acquisition window is determined, and the first acquisition window includes a first rotation angle; based on the laser detection path and laser detection parameters of the laser emission device, a first light spot coverage area is determined, and the first light spot coverage area includes a second rotation angle; according to the synchronous control gate circuit, in combination with the first acquisition window and the first light spot coverage area, the monocular camera and the laser emission device are linked and controlled.

[0085] Further, the visual feature point extraction module 10 of the embodiment of the present application is further configured to perform the following steps:

[0086] Monitor the change of ambient light in real time to obtain the external light conditions; at the same time, evaluate the clarity according to the point cloud data detected by the laser emission device and the visual feature points associated with the monocular camera, where the clarity includes the density of the visual feature points and the density of the point cloud data; based on the external light conditions, in combination with the clarity, dynamically adjust the exposure time of the monocular camera and the laser intensity of the laser emission device.

[0087] Further, the visual feature point extraction module 10 of the embodiment of the present application is further configured to perform the following steps:

[0088] Obtain a first distribution sparse partition and a first distribution dense partition through the density of the visual feature points in the clarity; obtain a second distribution sparse partition and a second distribution dense partition through the density of the point cloud data in the clarity; based on the layout information of the transmission line, obtain the line key nodes, and the line key nodes are obtained by positioning with spatial complexity constraints; based on the first distribution sparse partition, the first distribution dense partition, the second distribution sparse partition, and the second distribution dense partition, in comparison with the line key nodes, add clarity constraint intervals.

[0089] Further, the device of the embodiment of the present application is further configured to perform the following steps:

[0090] Based on the layout information of the transmission line, perform distribution coordinate marking on each monocular camera to obtain a monocular camera distribution mark; according to the monocular camera distribution mark, in comparison with the laser detection path, determine a monocular camera collaborative activation sequence; based on the monocular camera collaborative activation sequence and the monocular camera distribution mark, match a monocular camera cross-activation sequence, and the monocular camera collaborative activation sequence corresponds to the monocular camera cross-activation sequence one by one.

[0091] Further, the device of the embodiment of the present application is further configured to perform the following steps:

[0092] Generate a scanning coverage space based on the laser detection path and in combination with the second rotation angle of the first light spot coverage area; in the scanning coverage space, locate the first collaborative monocular camera in the monocular camera collaborative activation sequence in combination with the first rotation angle of the first acquisition window; use the first collaborative monocular camera as the center, and determine a plurality of first cross monocular cameras according to the monocular camera distribution mark, where the first cross monocular cameras are located with time complexity constraints; traverse the monocular camera collaborative activation sequence according to the first collaborative monocular camera and the plurality of first cross monocular cameras to determine the monocular camera cross activation sequence.

[0093] Further, the relative space position limiting module 40 in the embodiments of the present application is further configured to perform the following steps:

[0094] In the scanning coverage space, set eight quadrant sub-spaces by constraining the cross-space phase of the cross-line laser; based on the eight quadrant sub-spaces and the first collaborative monocular camera, perform relative position encoding on a plurality of first cross monocular cameras, where the first collaborative monocular camera is assigned a coding value according to the position relationship of the eight quadrant sub-spaces; based on the first collaborative monocular camera and the plurality of first cross monocular cameras, establish a spatial index for the three-dimensional reconstruction of the transmission line through the relative position encoding.

[0095] Through the foregoing detailed description of the method for three-dimensional reconstruction of a transmission line using a cross laser in this specification, those skilled in the art can clearly know the device for three-dimensional reconstruction of a transmission line using a cross laser in this embodiment. For the device disclosed in Embodiment 2, since it corresponds to the method disclosed in Embodiment 1, it has corresponding functional modules and beneficial effects. For the relevant parts, reference may be made to the description in the method section.

[0096] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for three-dimensional reconstruction of power transmission lines using cross laser, characterized in that: The method comprises: Through the linkage control of the monocular camera and the laser emission equipment, the cross spot formed on the transmission line and the surrounding environment is used to extract the visual feature points; At the same time, the trajectory coordinates of the monocular camera during shooting are recorded to obtain the external parameters and internal parameters of the monocular camera; Based on the layout information of the transmission line, configure the laser detection path and laser detection parameters; Determining the positional relationship between the cross light spot and the transmission line, and limiting the relative spatial position between the transmission channels by constraining the cross spatial phase of the cross line laser, including: In the scanning coverage space, eight quadrant subspaces are set by constraining the cross-space phase of the cross-line laser; Based on the eight quadrant subspaces and the first cooperative monocular camera, a plurality of first cross monocular cameras are relatively position-encoded, wherein the first cooperative monocular camera is assigned a coding value according to a position relationship of the eight quadrant subspaces; Based on the first cooperative monocular camera and the plurality of first crossed monocular cameras, a spatial index for three-dimensional reconstruction of the power transmission line is established through the relative position encoding; The point cloud data detected by the laser emitting device and the visual feature points associated with the monocular camera are fused, and the external parameters and internal parameters of the monocular camera, the laser detection path and the laser detection parameters are combined to perform three-dimensional reconstruction of the transmission line under the constraint of the cross space phase.

2. The method for three-dimensional reconstruction of power transmission lines using cross laser as claimed in claim 1, characterized in that: Through the linkage control of a monocular camera and a laser emission device, a cross spot is formed on the power transmission line and the surrounding environment to extract visual feature points. The method includes: Filter the cross light spots formed on the transmission lines and the surrounding environment to obtain the cross light spot feature points; Based on the cross light spot feature points, a light spot classification model is trained, where the light spot classification model is used to filter non-target light spots; The light spot classification model is used to retain the target light spot and extract the visualization feature points.

3. The method for three-dimensional reconstruction of power transmission lines using cross laser as claimed in claim 2, characterized in that: The monocular camera and the laser emitting device are controlled in linkage to synchronize the timing of the synchronous control gate circuit; Determining a first acquisition window based on external parameters and internal parameters of the monocular camera, wherein the first acquisition window includes a first rotation angle; The first rotation angle is the rotation angle of the monocular camera when capturing images, and is used to determine the range of image capture; Determine a first light spot coverage area based on a laser detection path and laser detection parameters of the laser emitting device, wherein the first light spot coverage area includes a second rotation angle; The second rotation angle is the rotation angle of the laser emitting device when emitting laser, and is used to determine the range covered by the light spot; According to the synchronous control gate circuit, in combination with the first acquisition window and the first light spot coverage area, the monocular camera and the laser emitting device are controlled in linkage.

4. The method for three-dimensional reconstruction of power transmission lines using cross laser as claimed in claim 3, characterized in that: According to the synchronous control gate circuit, in combination with the first acquisition window and the first light spot coverage area, the monocular camera and the laser emitting device are controlled in linkage, and the method includes: Monitor ambient light changes in real time and obtain external lighting conditions; At the same time, according to the point cloud data detected by the laser emitting device and the visual feature points associated with the monocular camera, the clarity is evaluated, and the clarity includes the density of the visual feature points and the density of the point cloud data; Based on the external lighting conditions and in combination with the clarity, the exposure time of the monocular camera and the laser intensity of the laser emitting device are dynamically adjusted.

5. The method for three-dimensional reconstruction of power transmission lines using cross laser as claimed in claim 4, characterized in that: Based on the external lighting conditions and in combination with the clarity, the exposure time of the monocular camera and the laser intensity of the laser emitting device are dynamically adjusted, and the method includes: Obtaining a first distribution sparse partition and a first distribution dense partition according to the density of the visualized feature points in the clarity; Obtaining a second distribution sparse partition and a second distribution dense partition according to the density of the point cloud data in the definition; Based on the layout information of the transmission line, key nodes of the line are obtained, and the key nodes of the line are located according to the spatial complexity constraint; Based on the first distribution sparse partition, the first distribution dense partition, the second distribution sparse partition, and the second distribution dense partition, clarity constraint intervals are added in comparison with the line key nodes.

6. The method for three-dimensional reconstruction of power transmission lines using cross laser as claimed in claim 5, characterized in that: Based on the layout information of the transmission line, the distribution coordinates of each monocular camera are marked to obtain the monocular camera distribution mark; Determine a monocular camera collaborative activation sequence according to the monocular camera distribution mark and the laser detection path; Based on the monocular camera collaborative activation sequence and the monocular camera distribution mark, the monocular camera cross activation sequence is matched, and the monocular camera collaborative activation sequence corresponds to the monocular camera cross activation sequence one by one.

7. The method for three-dimensional reconstruction of power transmission lines using cross laser as claimed in claim 6, characterized in that: The monocular camera collaborative activation sequence corresponds one-to-one to the monocular camera cross activation sequence, and the method includes: Based on the laser detection path, combined with the second rotation angle of the first light spot coverage area, a scanning coverage space is generated; In the scanning coverage space, in combination with a first rotation angle of the first acquisition window, positioning a first cooperative monocular camera in the monocular camera cooperative activation sequence; Taking the first cooperative monocular camera as the center, and determining a plurality of first cross monocular cameras according to the monocular camera distribution marks, wherein the first cross monocular cameras are located according to time complexity constraints; According to the first cooperative monocular camera and a plurality of first cross monocular cameras, the monocular camera cooperative activation sequence is traversed to determine the monocular camera cross activation sequence.

8. A device for three-dimensional reconstruction of power transmission lines using cross laser, characterized in that: The device is used to execute the method for three-dimensional reconstruction of a power transmission line by cross laser according to any one of claims 1 to 7, and the device comprises: A visualization feature point extraction module, which is used to extract visualization feature points from a cross spot formed on the power transmission line and the surrounding environment through linkage control of a monocular camera and a laser emission device; A monocular camera parameter acquisition module, which is used to record the trajectory coordinates of the monocular camera during shooting, and obtain the external parameters and internal parameters of the monocular camera; A laser detection configuration module, the laser detection configuration module is used to configure a laser detection path and laser detection parameters based on the layout information of the transmission line; A relative spatial position limiting module, which is used to determine the positional relationship between the cross light spot and the power transmission line, and to limit the relative spatial position between the power transmission channels by constraining the cross spatial phase of the cross-line laser; A three-dimensional reconstruction module is used to fuse the point cloud data detected by the laser emitting device and the visual feature points associated with the monocular camera, and to perform three-dimensional reconstruction of the transmission line under the constraint of the cross space phase by combining the external parameters and internal parameters of the monocular camera, the laser detection path and the laser detection parameters.

Images