Intelligent management method and system for electric power construction progress

Abstract

The invention discloses a power construction progress intelligent management method and system, and the method comprises the steps: carrying out the periodic cruise according to a preset height measurement position through an unmanned plane, and collecting a video stream of iron tower construction in real time; identifying the current construction height based on the video stream, comparing the current construction height with a preset height threshold value, judging that a construction progress node is reached when any threshold value is reached, and generating a trigger signal; responding to the signals to obtain coordinates of corresponding installation monitoring points, and controlling the unmanned aerial vehicle to collect videos at fixed points; component installation state analysis is carried out on the fixed-point video, and an abnormal alarm is generated when the feature conformity is lower than a threshold value; marking the timestamps of the video clips corresponding to the thresholds, and associatively storing the timestamps to a construction progress node; and generating a visual progress interface based on the association result. Through unmanned aerial vehicle cruise acquisition, video analysis and data association technologies, quantitative management of the construction progress, accurate detection of the installation quality, full-dimension tracing of the process and safety risk prevention and control are realized, and the construction efficiency and safety of the electric power iron tower are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of machine vision detection, and in particular to an intelligent management method and system for electric power construction progress. Background Art

[0002] As critical infrastructure for power transmission lines, the construction quality of power towers is directly impacted by the safety and reliability of grid operations. Traditionally, the manual modular assembly of power towers involves workers assembling components, such as angle steel, connecting plates, and bolts, layer by layer, starting from the tower base and working upwards. This process relies on manual visual inspection and empirical judgment, presenting significant technical bottlenecks. First, working at height makes it difficult for workers to comprehensively and accurately verify the installation status of each component, leading to potential issues such as missing bolts, missing gaskets, and misaligned components. These risks can gradually manifest over long-term operation due to factors such as load fluctuations and environmental erosion, ultimately leading to safety incidents such as tower structural instability. Second, manual progress management, primarily through on-site measurements and paper records, can result in delayed progress assessments and difficulty tracing data, failing to meet the demands of modern power projects for construction efficiency and refined management. Summary of the Invention

[0003] In order to solve at least one of the technical problems mentioned above, the present invention provides an intelligent management method and system for power construction progress.

[0004] In a first aspect, the present invention provides a method for intelligent management of power construction progress, the method comprising: Control the drone to perform periodic patrols according to the preset tower height and collect real-time video streams of the tower construction; Identifying a current construction height of the tower based on the video stream, and comparing the current construction height with a plurality of preset height thresholds; When it is detected that the current construction height reaches any height threshold, it is determined that the construction progress node has been reached and a node trigger signal is generated; In response to the node trigger signal, the coordinates of the installation monitoring point corresponding to the height threshold are obtained, and the UAV is controlled to fly to the installation monitoring point for fixed-point video acquisition; Analyze component installation status in videos collected at fixed points. If the degree of conformity with preset features is detected to be lower than the threshold, an installation anomaly alarm is generated. Timestamp the video segments corresponding to each height threshold in the video stream and store them in association with the corresponding construction progress nodes; A visual progress interface is generated based on the association results, and the visual progress interface supports calling and playing associated video clips according to construction progress nodes.

[0005] Preferably, the identifying the current construction height of the tower based on the video stream includes: Perform multi-scale edge detection on key frames of video streams; use adaptive threshold segmentation to extract the main pole area; Probabilistic Hough transform is used to identify the straight lines in the main member area and calculate the endpoint coordinates of each straight line; Construct a vertical projection histogram based on the endpoint coordinate set; Determine the height corresponding to the maximum peak in the vertical projection histogram as the current construction height.

[0006] Preferably, obtaining the coordinates of the installation monitoring point corresponding to the height threshold includes: Pre-establish a database of coordinate mapping of monitoring points for height threshold space installation; When any height threshold is triggered, the coordinates of multiple installation monitoring points corresponding to the height threshold are queried through the hash table.

[0007] Preferably, the component installation status analysis of the video collected at the fixed point and the generation of an installation abnormality alarm if it is detected that the preset feature compliance is lower than a threshold value include: Use edge detection algorithm to extract component contour feature points; Perform Delaunay triangulation on multiple component contour feature points to generate a topology comparison mesh; and calculate the Hausdorff distance between the topology comparison mesh and the standard mesh; Convert the image to HSV space, extract the hue distribution histogram of the bolt connection area; calculate the Bhattacharyya distance between the hue distribution histogram and the standard sample; If the Bhattacharyya distance is lower than a preset threshold and the Hausdorff distance exceeds a preset pixel value, an installation abnormality alarm is generated.

[0008] Preferably, the associated storage to the corresponding construction progress node includes: Structured metadata is added to each video clip, where the structured metadata includes the construction stage number, height threshold, acquisition time, drone positioning coordinates, and environmental parameters.

[0009] In a second aspect, the present invention further provides an intelligent management system for electric power construction progress, the system comprising: The UAV cruise acquisition module is used to control the UAV to perform periodic cruises according to the preset tower height measurement position and collect real-time video streams of the tower construction; A construction height identification module is used to identify the current construction height of the tower based on the video stream and compare the current construction height with a plurality of preset height thresholds; The progress node determination module is used to determine that the construction progress node has been reached when it is detected that the current construction height has reached any height threshold, and to generate a node trigger signal; A fixed-point monitoring control module is configured to respond to the node trigger signal, obtain the coordinates of the installation monitoring point corresponding to the height threshold, and control the drone to fly to the installation monitoring point for fixed-point video acquisition; The component installation status analysis module is used to analyze the component installation status of the video collected at the fixed point. If the preset feature compliance is detected to be lower than the threshold, an installation abnormality alarm is generated; A video segment management module is used to timestamp the video segments corresponding to each height threshold in the video stream and associate and store them with the corresponding construction progress nodes; The visual progress display module is used to generate a visual progress interface based on the association results. The visual progress interface supports calling and playing associated video clips according to the construction progress nodes.

[0010] Preferably, the construction height identification module is further used to: Perform multi-scale edge detection on key frames of video streams; use adaptive threshold segmentation to extract the main pole area; Probabilistic Hough transform is used to identify the straight lines in the main member area and calculate the endpoint coordinates of each straight line; Construct a vertical projection histogram based on the endpoint coordinate set; Determine the height corresponding to the maximum peak in the vertical projection histogram as the current construction height.

[0011] Preferably, the component installation status analysis module is further used to: Use edge detection algorithm to extract component contour feature points; Perform Delaunay triangulation on multiple component contour feature points to generate a topology comparison mesh; and calculate the Hausdorff distance between the topology comparison mesh and the standard mesh; Convert the image to HSV space, extract the hue distribution histogram of the bolt connection area; calculate the Bhattacharyya distance between the hue distribution histogram and the standard sample; If the Bhattacharyya distance is lower than a preset threshold and the Hausdorff distance exceeds a preset pixel value, an installation abnormality alarm is generated.

[0012] In a third aspect, the present invention also provides an electronic device comprising a processor and a memory, wherein the memory is used to store computer program code, and the computer program code comprises computer instructions. When the processor executes the computer instructions, the electronic device executes the method as described in the first aspect above and any possible implementation thereof.

[0013] In a fourth aspect, the present invention also provides a computer-readable storage medium, in which a computer program is stored. The computer program includes program instructions, and when the program instructions are executed by a processor of an electronic device, the processor executes the method as described in the first aspect above and any possible implementation method thereof.

[0014] Compared with the prior art, the present invention has the following beneficial effects: 1) The present invention uses drone cruising to collect video streams, identify construction heights, and trigger node detection. It then combines fixed-point video analysis to analyze component installation status, mark and store related video clips, and generate a visual interface. This enables intelligent quantitative management of construction progress, precise detection of installation quality, and full-dimensional traceability of the construction process. This improves management efficiency and proactively prevents and controls safety risks, effectively resolving issues such as uncontrolled progress and quality risks in manual assembly, and significantly enhancing construction safety and management efficiency.

[0015] 2) The present invention extracts component contour feature points through edge detection and performs Delaunay triangulation. After generating a topological structure comparison grid, the Hausdorff distance with the standard grid is calculated to detect geometric layout deviations. At the same time, the image is converted to HSV space to extract the hue distribution histogram of the bolt connection area and the Bhattacharyya distance with the standard sample is calculated to identify material color differences. When the Bhattacharyya distance is lower than the preset threshold and the Hausdorff distance exceeds the preset pixel value, an installation abnormality alarm is generated, realizing multimodal fusion detection of geometric structure and color features. It not only ensures that the judgment standard is consistent with the engineering specification through quantitative indicators, but also utilizes the illumination robustness of HSV space and the structural stability of triangulation to improve the anti-interference ability in complex environments. It can also accurately distinguish different types of installation defects through dual-indicator combination logic, providing clear guidance for construction rectification, effectively solving the problems of high missed judgment rate and low standardization in traditional manual detection, and significantly improving the accuracy and reliability of power tower installation quality inspection.

[0016] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the background technology, the drawings required for use in the embodiments of the present invention or the background technology will be described below.

[0018] The accompanying drawings herein are incorporated into and constitute a part of the specification. These drawings illustrate embodiments consistent with the present disclosure and, together with the specification, are used to explain the technical solutions of the present disclosure.

[0019] Figure 1 A schematic diagram of a flow chart of an intelligent management method for electric power construction progress provided by an embodiment of the present invention; Figure 2 A schematic diagram of the structure of an intelligent management system for power construction progress provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0021] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present invention. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute a separate or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0022] Traditional manual assembly of power towers has problems with incomplete installation or missing installation, and manual progress management is inefficient. Existing drone monitoring lacks the ability to conduct refined inspections of key components, automated defect identification, and structured data management of the construction process, resulting in a high missed detection rate, poor environmental adaptability, and difficulty in traceability.

[0023] See also Figure 1 , Figure 1 The following is a flow chart of an intelligent management method for power construction progress provided by an embodiment of the present invention. Figure 1 As shown, the method includes: S100 controls the drone to periodically patrol according to the preset tower height and collect real-time video streams of the tower construction; The drone is controlled to perform periodic patrols based on pre-defined tower height measurement locations. Specifically, the drone's built-in global positioning system (GPS) and inertial navigation system work together to plan multiple circular routes within the tower construction area. Each route includes multiple equally spaced height measurement locations, corresponding to typical height levels during different tower construction stages (e.g., foundation, standard, and necking). The drone is equipped with a high-definition camera to capture real-time video streams of the tower construction. This video stream is embedded with metadata such as the drone's real-time position coordinates, flight altitude, and heading angle, generating dynamic monitoring data with spatial and temporal information. The pre-defined tower height measurement locations are pre-calibrated 3D coordinates based on the theoretical heights of each structural layer as specified in the tower design drawings and the construction site's geographic elevation data. This ensures that the drone's patrols cover the entire tower construction height range and key structural areas.

[0024] S200, identifying a current construction height of the tower based on the video stream, and comparing the current construction height with a plurality of preset height thresholds; In this embodiment, the contour features of the tower components are extracted using the Canny operator edge detection algorithm. A scale-invariant feature transformation algorithm is then used to perform feature matching on landmark components such as angle steel and connecting plates in adjacent frame images. Combined with the drone's real-time flight altitude data, the top height of the currently installed tower structure is calculated using the principle of triangulation. Multiple preset height thresholds are set based on the design heights of key construction nodes in the tower construction drawings. For example, a height threshold is set for each completed 5-meter standard section, including the completed height of the foundation layer, the height of the standard layer segment, and the height of the tower top lightning rod installation. The current construction height calculated in real time is numerically compared with the aforementioned thresholds to determine whether to trigger the subsequent node detection process.

[0025] S300, when it is detected that the current construction height reaches any height threshold, it is determined that a construction progress node has been reached and a node trigger signal is generated; When the current construction height reaches any threshold, the system determines that a construction progress node has been reached. If the difference between the current construction height and a preset threshold is within the allowable error range (e.g., ±0.5 meters), construction at that height level is considered complete, and a node trigger signal is generated containing information such as the node number, preset height, and actual arrival time. For example, when the tower top reaches 30 meters (corresponding to the preset threshold for completion of the third standard section), the node is marked as "Standard Section Construction Completed - Section 3" and a trigger command is sent to subsequent modules, ensuring quantitative management and node-based control of the construction progress.

[0026] S400, in response to the node trigger signal, obtaining the coordinates of the installation monitoring point corresponding to the height threshold, and controlling the drone to fly to the installation monitoring point to perform fixed-point video acquisition; In response to a node trigger signal, the coordinates of the installation monitoring point corresponding to that height threshold are obtained. This is achieved through a pre-established construction node-monitoring point coordinate mapping database. Specifically, each height threshold is associated with a set of three-dimensional coordinates corresponding to the installation locations of key components at that height level of the tower (such as the connection points between the crossarm and the tower body, the distribution of bolt groups, and the suspension points of insulator strings). When controlling the drone to fly to the installation monitoring point for fixed-point video capture, the drone first switches to fixed-point hovering mode, adjusts the camera's pitch angle and focal length, and captures close-up images of the monitoring point, providing a high-definition data foundation for subsequent installation status analysis.

[0027] S500: Analyze the component installation status of the video collected at the fixed point. If it is detected that the preset feature compliance is lower than the threshold, an installation abnormality alarm is generated; Component installation status analysis is performed on fixed-point video. Using deep learning-based object detection and semantic segmentation algorithms, the system detects components such as angle steel, bolts, washers, and connecting plates within the video frames, extracting their location coordinates, dimensional parameters, and connection status characteristics. The system then calculates the conformance of each component to its pre-set characteristics using a pre-set installation specification knowledge base (including rules such as component spacing, bolt torque standards, and anti-loosening washer installation angles). For example, the exposed thread length of a bolt should be no less than 2 times the thread pitch. If a bolt is detected with an exposed length of only 1 times the thread pitch, the conformance is calculated as 50%. If the conformance of any key component's pre-set characteristics falls below a pre-set threshold (e.g., 80%), an installation anomaly alert is generated, including the abnormal location coordinates, component type, and a description of the non-conformance. This alert is sent to the on-site construction manager via SMS and a system pop-up window, enabling accurate identification and real-time warning of issues such as missing or inadequate installation.

[0028] S600: Timestamp the video segments corresponding to the height thresholds in the video stream and store them in association with the corresponding construction progress nodes; The video segments corresponding to each height threshold in the video stream are timestamped. When a node trigger signal is detected, 5 minutes of video segments before and after the node are captured from the video stream (covering the construction process before and after the node arrives). Timestamps accurate to milliseconds are added to the segments. The timestamps contain year, month, day, hour, minute, second, and millisecond information to ensure the uniqueness and traceability of the time record. When associated and stored with the corresponding construction progress node, a three-dimensional index of "construction progress node-timestamp-video segment path" is established through a relational database. Each node can be associated with multiple video segments of different time periods. For example, the foundation layer node can be associated with videos of the concrete pouring process, the anchor bolt installation process, the first-floor angle steel assembly process, etc., forming a three-dimensional construction process data archive, providing complete data support for subsequent construction review and responsibility tracing.

[0029] S700: Generate a visual progress interface based on the correlation result, wherein the visual progress interface supports retrieving and playing associated video clips according to the construction progress nodes.

[0030] A 3D visualization model was constructed using the front-end development framework Vue.js, showcasing the real-time progress of tower construction in a three-dimensional view within the interface. Each construction progress node corresponds to an interactive icon within the model; clicking on an icon retrieves and plays the associated video clip. The interface supports navigating by construction progress node. Users can quickly locate the construction phase corresponding to a certain height threshold by sliding the timeline or filtering the node list. The associated video clips are retrieved from the database and played continuously in chronological order within the player, while information such as the theoretical construction time, actual completion time, and test results for that node are overlaid. For example, when a user clicks on a "crossarm installation node," a video of the installation process for that node pops up on the right side of the interface, while a bolt inspection report and abnormality alarm records are displayed simultaneously below, enabling transparent management of construction progress and multi-dimensional data linkage.

[0031] In this embodiment, drone cruise is used to collect video streams, identify construction heights and trigger node detection. Combined with fixed-point video analysis to analyze component installation status, mark and store related video clips, and generate a visual interface, it realizes intelligent quantitative management of construction progress, precise detection of installation quality, and full-dimensional traceability of the construction process, improves management efficiency and proactively prevents and controls safety risks, effectively solves the problems of progress out of control and quality risks in manual assembly, and significantly improves construction safety level and management efficiency.

[0032] Preferably, the identifying the current construction height of the tower based on the video stream includes: Perform multi-scale edge detection on key frames of video streams; use adaptive threshold segmentation to extract the main pole area; Probabilistic Hough transform is used to identify the straight lines in the main member area and calculate the endpoint coordinates of each straight line; Construct a vertical projection histogram based on the endpoint coordinate set; Determine the height corresponding to the maximum peak in the vertical projection histogram as the current construction height.

[0033] Multi-scale edge detection is performed on key frames of the video stream. This involves smoothing single frames (key frames) extracted every one second from the video stream using Gaussian filters with varying standard deviations, such as σ = 1.0, σ = 1.5, and σ = 2.0, to suppress noise at different scales. The Canny edge detection algorithm is then used to extract edge contours at each scale, forming a multi-scale edge image set. This process effectively captures the edge features of the tower's main poles under varying resolutions and lighting conditions. Adaptive threshold segmentation is used to extract the main pole regions. Grayscale analysis of the multi-scale edge images is performed using the Otsu algorithm, and an optimal threshold is calculated to distinguish the tower's main poles (supporting structures such as angle steel and channel steel) from the background (sky, ground, construction equipment, etc.). A binary image is generated, with the white area representing the extracted main pole regions. This adaptive processing avoids the segmentation bias that can occur with fixed thresholds in complex environments. Furthermore, a probabilistic Hough transform (HOT) algorithm is used to identify straight lines within the main pole region. This algorithm, based on the traditional Hough transform, introduces a random sampling mechanism, performing probabilistic voting only on edge points within the main pole region. By setting an accumulator threshold (e.g., 50 edge point votes) and a line segment length threshold (e.g., 50 pixels), it rapidly detects continuous and meaningful straight line segments (e.g., the edges of the tower's vertical main limbs and the connecting lines of the horizontal crossbars). The endpoint coordinates (x_i, y_i) of each straight line are then calculated, where the y coordinate corresponds to the vertical height position in the image. This method reduces computational complexity by over 90% compared to the traditional Hough transform, improving detection efficiency. A vertical projection histogram is constructed based on the endpoint coordinate set. This involves counting the y coordinates of all endpoints, dividing them into bins of 1 pixel, and calculating the frequency of endpoint occurrence within each vertical height bin. This results in a histogram with frequency as the y-axis and vertical height as the x-axis. For example, when the top endpoints of a tower's main pole appear concentrated in the image pixel height interval of 1200-1210, the frequency value in this interval is significantly higher than that in the lower interval. Furthermore, the height corresponding to the maximum peak in the vertical projection histogram is determined to be the current construction height. That is, by traversing the histogram to find the vertical height interval with the highest frequency, the midpoint value of the interval is taken as the current construction height. Combining the UAV flight altitude and the camera calibration parameters (such as focal length and pitch angle), the pixel height is converted into the actual physical height (unit: meter). For example, when the camera calibration coefficient is 0.05 meters / pixel, a pixel height of 1200 corresponds to an actual height of 60 meters.

[0034] In this embodiment, multi-scale edge detection is used to adapt to the complex construction environment lighting and background interference, combined with adaptive threshold segmentation to accurately extract the main pole area, and the probabilistic Hough transform is used to efficiently identify long straight line segments and calculate the endpoint coordinates. The construction height is then determined by the peak statistics of the vertical projection histogram. This not only improves the recognition robustness and detection efficiency of the main structure of the tower under complex conditions, but also enhances the anti-noise ability through statistical features, solves the subjective errors and insufficient environmental adaptability of traditional manual measurement, and provides high-precision and high-reliability key technical support for intelligent management of construction progress.

[0035] Preferably, obtaining the coordinates of the installation monitoring point corresponding to the height threshold includes: Pre-establish a database of coordinate mapping of monitoring points for height threshold space installation; When any height threshold is triggered, the coordinates of multiple installation monitoring points corresponding to the height threshold are queried through the hash table.

[0036] A database of spatially mapped monitoring points for height thresholds is pre-built using Building Information Modeling (BIM) technology. This database is based on the 3D structural model in the tower design drawings and incorporates the requirements for component connection strength and installation accuracy in power construction specifications. For each preset height threshold (such as the finished foundation layer height, standard section layer height, and tower top equipment installation height), several key installation locations are calibrated in 3D space as monitoring points. The coordinates of each monitoring point include X, Y, and Z spatial parameters. The Z axis corresponds to the tower's height, while the X and Y coordinates are determined based on the tower's cross-sectional dimensions (such as the side length of square or rectangular towers) and component distribution patterns (such as the crossarm-to-tower connection point, flange bolt group center, and insulator string suspension support point). For example, for a standard section mid-level with a height threshold of 30 meters, the database pre-stores the coordinates of four tower limb connection points, two crossarm support angle steel installation points, and one horizontal partition fixing point, forming a set of monitoring points that correspond to each height threshold. The database is built using a relational database management system (such as PostgreSQL) and includes a "height threshold table" and a "monitoring point coordinate table," with data mapping achieved through foreign key associations. When any height threshold is triggered, a hash table is used to query the coordinates of multiple installed monitoring points corresponding to the corresponding height threshold. Specifically, a hash data structure (such as a chained hash table) is used to store the mapping relationship between the height threshold and the set of monitoring point coordinates. The height threshold value (such as 30.0 meters) is used as the hash key. A hash function (such as the MD5 hash algorithm) converts the key value into a fixed-length index value, allowing for rapid location of the memory address storing the monitoring point coordinates. For example, when step S300 detects that the current construction height reaches the preset threshold of 20 meters, the hash table immediately returns the coordinates of the five monitoring points corresponding to that threshold (including the location of the bolt group of the connecting plate in the middle of the tower body, the locations of the welding points between the diagonal member and the main limb, etc.).

[0037] Preferably, the component installation status analysis of the video collected at the fixed point and the generation of an installation abnormality alarm if it is detected that the preset feature compliance is lower than a threshold value include: Use edge detection algorithm to extract component contour feature points; Perform Delaunay triangulation on multiple component contour feature points to generate a topology comparison mesh; and calculate the Hausdorff distance between the topology comparison mesh and the standard mesh; Convert the image to HSV space, extract the hue distribution histogram of the bolt connection area; calculate the Bhattacharyya distance between the hue distribution histogram and the standard sample; If the Bhattacharyya distance is lower than a preset threshold and the Hausdorff distance exceeds a preset pixel value, an installation abnormality alarm is generated.

[0038] An edge detection algorithm is used to extract component contour feature points. Specifically, a single-frame image of a fixed-point video is preprocessed, and after suppressing noise with a Gaussian filter (standard deviation σ=1.2), the Canny edge detection algorithm is applied. By calculating the gradient amplitude and direction, non-maximum suppression, and double-threshold hysteresis connection, the closed contour edges of components such as angle steel, bolts, and connecting plates are extracted. Feature points (x_i, y_i) are uniformly sampled at intervals of 5 pixels on the contour to form a discrete point set containing component geometric shape information. Delaunay triangulation is performed on multiple component contour feature points. Based on the empty circle characteristics, the feature points are connected into non-overlapping triangular meshes to generate a topological structure comparison mesh that reflects the relative positional relationship of the components. For example, for the bolt group area, the triangulation can present the spatial layout of the bolt hole and the edge of the connecting plate. The geometric shape difference between the two is quantified by calculating the Hausdorff distance between the topological structure comparison mesh and the standard mesh (based on the ideal component layout preset in the design drawing). The Hausdorff distance is defined as the maximum one-way distance between two point sets, and the calculation formula is: When bolts are missing or misaligned, this distance value increases significantly. Furthermore, the image is converted to the HSV color space (H represents hue, S represents saturation, and V represents value). A hue distribution histogram is extracted for the bolted joint area (located through preliminary target detection). The hue dimension (0°-360°) is divided into 32 bins, and the frequency of pixels within each bin is counted to form a probability distribution vector. The Bhattacharyya distance between this histogram and a standard sample (pre-stored standard color distributions for bolts, washers, and anti-corrosion coatings) is calculated to measure the similarity between the two probability distributions. When bolts are not installed with anti-loosening washers according to specifications or the coating color is abnormal, the Bhattacharyya distance value will fall below a preset threshold (e.g., 0.3). If the Bhattacharyya distance falls below the preset threshold and the Hausdorff distance exceeds a preset pixel value (e.g., 15 pixels), an installation anomaly is determined, and an alert message is generated, including the anomaly type (e.g., missing bolts, misaligned components), location coordinates, and image evidence. For example, when it is detected that the Hausdorff distance between the triangulation grid of a certain connecting plate area and the standard grid is 20 pixels (exceeding the threshold of 15 pixels), and the Bhattacharyya distance of the hue histogram of the bolt hole area is 0.25 (below the threshold of 0.3), it is identified as "bolts missing and anti-loosening gaskets are missing", and an alarm is immediately pushed to the construction management platform.

[0039] In this embodiment, edge detection is used to extract component contour feature points and Delaunay triangulation is performed. After generating a topological structure comparison grid, the Hausdorff distance with the standard grid is calculated to detect geometric layout deviations. At the same time, the image is converted to HSV space to extract the hue distribution histogram of the bolt connection area and calculate the Bhattacharyya distance with the standard sample to identify material color differences. When the Bhattacharyya distance is lower than a preset threshold and the Hausdorff distance exceeds a preset pixel value, an installation abnormality alarm is generated, realizing multimodal fusion detection of geometric structure and color features. It not only ensures that the judgment standard is consistent with the engineering specification through quantitative indicators, but also utilizes the illumination robustness of the HSV space and the structural stability of the triangulation to improve the anti-interference ability in complex environments. It can also accurately distinguish different types of installation defects through dual-indicator combination logic, providing clear guidance for construction rectification, effectively solving the problems of high missed judgment rate and low standardization in traditional manual inspection, and significantly improving the accuracy and reliability of power tower installation quality inspection.

[0040] Preferably, the associated storage to the corresponding construction progress node includes: Structured metadata is added to each video clip, where the structured metadata includes the construction stage number, height threshold, acquisition time, drone positioning coordinates, and environmental parameters.

[0041] In summary, the method provided in this embodiment can at least achieve the following effects: 1) The present invention uses drone cruising to collect video streams, identify construction heights, and trigger node detection. It then combines fixed-point video analysis to analyze component installation status, mark and store related video clips, and generate a visual interface. This enables intelligent quantitative management of construction progress, precise detection of installation quality, and full-dimensional traceability of the construction process. This improves management efficiency and proactively prevents and controls safety risks, effectively resolving issues such as uncontrolled progress and quality risks in manual assembly, and significantly enhancing construction safety and management efficiency.

[0042] 2) The present invention extracts component contour feature points through edge detection and performs Delaunay triangulation. After generating a topological structure comparison grid, the Hausdorff distance with the standard grid is calculated to detect geometric layout deviations. At the same time, the image is converted to HSV space to extract the hue distribution histogram of the bolt connection area and the Bhattacharyya distance with the standard sample is calculated to identify material color differences. When the Bhattacharyya distance is lower than the preset threshold and the Hausdorff distance exceeds the preset pixel value, an installation abnormality alarm is generated, realizing multimodal fusion detection of geometric structure and color features. It not only ensures that the judgment standard is consistent with the engineering specification through quantitative indicators, but also utilizes the illumination robustness of HSV space and the structural stability of triangulation to improve the anti-interference ability in complex environments. It can also accurately distinguish different types of installation defects through dual-indicator combination logic, providing clear guidance for construction rectification, effectively solving the problems of high missed judgment rate and low standardization in traditional manual detection, and significantly improving the accuracy and reliability of power tower installation quality inspection.

[0043] See also Figure 2 In one embodiment, an intelligent management system for electric power construction progress is also provided, the system comprising: The drone cruise collection module 100 is used to control the drone to perform periodic cruises according to the preset tower height measurement position and collect the video stream of the tower construction in real time; A construction height identification module 200 is configured to identify the current construction height of the tower based on the video stream and compare the current construction height with a plurality of preset height thresholds; The progress node determination module 300 is used to determine that a construction progress node has been reached when it is detected that the current construction height has reached any height threshold, and to generate a node trigger signal; The fixed-point monitoring control module 400 is configured to respond to the node trigger signal, obtain the coordinates of the installation monitoring point corresponding to the height threshold, and control the drone to fly to the installation monitoring point to perform fixed-point video acquisition; The component installation status analysis module 500 is used to analyze the component installation status of the video collected at the fixed point, and generate an installation abnormality alarm if it is detected that the preset feature compliance is lower than the threshold; The video segment management module 600 is used to timestamp the video segments corresponding to each height threshold in the video stream and associate and store them with the corresponding construction progress nodes; The visual progress display module 700 is used to generate a visual progress interface based on the correlation results. The visual progress interface supports calling and playing related video clips according to the construction progress nodes.

[0044] Preferably, the construction height identification module 200 is further used for: Perform multi-scale edge detection on key frames of video streams; use adaptive threshold segmentation to extract the main pole area; Probabilistic Hough transform is used to identify the straight lines in the main member area and calculate the endpoint coordinates of each straight line; Construct a vertical projection histogram based on the endpoint coordinate set; Determine the height corresponding to the maximum peak in the vertical projection histogram as the current construction height.

[0045] Preferably, the component installation status analysis module 500 is further used to: Use edge detection algorithm to extract component contour feature points; Perform Delaunay triangulation on multiple component contour feature points to generate a topology comparison mesh; and calculate the Hausdorff distance between the topology comparison mesh and the standard mesh; Convert the image to HSV space, extract the hue distribution histogram of the bolt connection area; calculate the Bhattacharyya distance between the hue distribution histogram and the standard sample; If the Bhattacharyya distance is lower than a preset threshold and the Hausdorff distance exceeds a preset pixel value, an installation abnormality alarm is generated.

[0046] It can be understood that the functions or modules included in the system provided in this embodiment can be used to execute the method described in the above method embodiment. Its specific implementation can refer to the description of the above method embodiment. For the sake of brevity, it will not be repeated here.

[0047] The present invention also provides an electronic device, including a processor and a memory, wherein the memory is used to store computer program code, and the computer program code includes computer instructions. When the processor executes the computer instructions, the electronic device executes a method as described in any one of the possible implementation modes.

[0048] The present invention also provides a computer-readable storage medium, in which a computer program is stored. The computer program includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor executes a method as described in any one of the possible implementation methods described above.

[0049] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0050] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here. Those skilled in the art will also clearly understand that the descriptions of the various embodiments of the present invention have different focuses. For the convenience and brevity of description, the same or similar parts may not be repeated in different embodiments. Therefore, for parts not described or not described in detail in a certain embodiment, reference can be made to the descriptions of other embodiments.

Claims

1. An intelligent management method for power construction progress, characterized in that: The method comprises: Control the drone to perform periodic patrols according to the preset tower height and collect real-time video streams of the tower construction; Identifying a current construction height of the tower based on the video stream, and comparing the current construction height with a plurality of preset height thresholds; When it is detected that the current construction height reaches any height threshold, it is determined that the construction progress node has been reached and a node trigger signal is generated; In response to the node trigger signal, the coordinates of the installation monitoring point corresponding to the height threshold are obtained, and the UAV is controlled to fly to the installation monitoring point for fixed-point video acquisition; Analyze component installation status in videos collected at fixed points. If the degree of conformity with preset features is detected to be lower than the threshold, an installation anomaly alarm is generated. Timestamp the video segments corresponding to each height threshold in the video stream and store them in association with the corresponding construction progress nodes; A visual progress interface is generated based on the association results, and the visual progress interface supports calling and playing associated video clips according to construction progress nodes.

2. The intelligent management method for electric power construction progress according to claim 1 is characterized in that: The identifying the current construction height of the tower based on the video stream includes: Perform multi-scale edge detection on key frames of video streams; use adaptive threshold segmentation to extract the main rod area; Probabilistic Hough transform is used to identify the straight lines in the main member area and calculate the endpoint coordinates of each straight line; Construct a vertical projection histogram based on the endpoint coordinate set; Determine the height corresponding to the maximum peak in the vertical projection histogram as the current construction height.

3. The intelligent management method for electric power construction progress according to claim 1 is characterized in that: The obtaining of the coordinates of the installation monitoring point corresponding to the height threshold includes: Pre-establish a database of coordinate mapping of monitoring points in height threshold space installation; When any height threshold is triggered, the coordinates of multiple installation monitoring points corresponding to the height threshold are queried through the hash table.

4. The intelligent management method for electric power construction progress according to claim 1 is characterized in that: The component installation status analysis is performed on the video collected at the fixed point. If it is detected that the preset feature compliance is lower than the threshold, an installation abnormality alarm is generated, including: Use edge detection algorithm to extract component contour feature points; Perform Delaunay triangulation on multiple component contour feature points to generate a topology comparison mesh; and calculate the Hausdorff distance between the topology comparison mesh and the standard mesh; Convert the image to HSV space, extract the hue distribution histogram of the bolt connection area; calculate the Bhattacharyya distance between the hue distribution histogram and the standard sample; If the Bhattacharyya distance is lower than a preset threshold and the Hausdorff distance exceeds a preset pixel value, an installation abnormality alarm is generated.

5. The intelligent management method for electric power construction progress according to claim 1 is characterized in that: The associated storage to the corresponding construction progress node includes: Structured metadata is added to each video clip, where the structured metadata includes the construction stage number, height threshold, acquisition time, drone positioning coordinates, and environmental parameters.

6. An intelligent management system for power construction progress, characterized in that: The system comprises: The UAV cruise acquisition module is used to control the UAV to perform periodic cruises according to the preset tower height measurement position and collect real-time video streams of the tower construction; A construction height identification module is used to identify the current construction height of the tower based on the video stream and compare the current construction height with a plurality of preset height thresholds; The progress node determination module is used to determine that the construction progress node has been reached when it is detected that the current construction height has reached any height threshold, and to generate a node trigger signal; A fixed-point monitoring control module is configured to respond to the node trigger signal, obtain the coordinates of the installation monitoring point corresponding to the height threshold, and control the drone to fly to the installation monitoring point for fixed-point video acquisition; The component installation status analysis module is used to analyze the component installation status of the video collected at the fixed point. If the preset feature compliance is detected to be lower than the threshold, an installation abnormality alarm is generated; A video segment management module is used to timestamp the video segments corresponding to each height threshold in the video stream and associate and store them with the corresponding construction progress nodes; The visual progress display module is used to generate a visual progress interface based on the association results. The visual progress interface supports calling and playing associated video clips according to the construction progress nodes.

7. The intelligent management system for electric power construction progress according to claim 6 is characterized in that: The construction height identification module is also used for: Perform multi-scale edge detection on key frames of video streams; use adaptive threshold segmentation to extract the main rod area; Probabilistic Hough transform is used to identify the straight lines in the main member area and calculate the endpoint coordinates of each straight line; Construct a vertical projection histogram based on the endpoint coordinate set; Determine the height corresponding to the maximum peak in the vertical projection histogram as the current construction height.

8. The intelligent management system for electric power construction progress according to claim 6 is characterized in that: The component installation status analysis module is also used for: Use edge detection algorithm to extract component contour feature points; Perform Delaunay triangulation on multiple component contour feature points to generate a topology comparison mesh; and calculate the Hausdorff distance between the topology comparison mesh and the standard mesh; Convert the image to HSV space and extract the hue distribution histogram of the bolt connection area; Calculate the Bhattacharyya distance between the hue distribution histogram and the standard sample; If the Bhattacharyya distance is lower than a preset threshold and the Hausdorff distance exceeds a preset pixel value, an installation abnormality alarm is generated.

9. An electronic device, characterized in that: include: A processor and a memory, the memory being used to store computer program code, the computer program code including computer instructions, and when the processor executes the computer instructions, the electronic device executes the intelligent management method for power construction progress as described in any one of claims 1 to 5.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor executes the intelligent management method for power construction progress as described in any one of claims 1 to 5.

Images