Bridge construction progress visual dynamic monitoring and early warning method and system based on unmanned aerial vehicle intelligent identification and BIM
Through the integration of intelligent drone identification and BIM technology, combined with AI algorithms and YOLOv5 network, real-time monitoring and early warning of bridge construction progress is achieved, and the problems of low construction progress monitoring efficiency and subjectivity of data analysis in the existing technology are solved, and the problems of low construction progress monitoring are improved. The level of intelligent construction management.
Patent Information
- Application Number
- CN202411019802.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-29
- Publication Date
- 2024-12-06
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The application of existing drone and artificial intelligence technologies in building construction management is mainly concentrated in a single field and has not been effectively integrated, resulting in inefficient construction progress monitoring and subjective and errors in data analysis.
Adopt a visual dynamic monitoring and early warning system for bridge construction progress based on intelligent drone identification and BIM, including model generation and viewing function modules, drone aerial data upload and viewing function modules, bridge component object detection function modules, and real-time monitoring and early warning The construction progress functional module optimizes the BIM model through AI algorithms and combines the YOLOv5 network for target detection to achieve real-time monitoring and early warning.
Real-time monitoring and intelligent management of bridge construction progress, improve the efficiency and accuracy of construction site data analysis, significantly improve the intelligence level of construction management, and provide convenient tools for remote monitoring and real-time data analysis.
Smart Images

Figure CN119090415A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of bridge construction progress monitoring, and in particular relates to a visual dynamic monitoring and early warning method and system for bridge construction progress based on unmanned aerial vehicle intelligent identification and BIM. Background Art
[0002] As construction projects continue to develop, monitoring construction progress is essential to ensure that projects are completed on time. Traditional methods of monitoring construction progress usually rely on manual inspections and data recording, which are inefficient and subjective and inaccurate. Combining drone aerial photography technology with artificial intelligence algorithms can achieve efficient monitoring and data analysis of construction sites and improve the level of intelligent construction management. The combination of digital technology, drone technology, and artificial intelligence technology has brought huge opportunities and challenges to construction management.
[0003] At present, the application of drones and artificial intelligence technology is mainly concentrated in a single technical field, such as image recognition, data analysis, etc. How to effectively integrate different technologies and apply them to actual construction management still requires further research and exploration.
[0004] Although some progress has been made in the application of drones and artificial intelligence technology in construction management at home and abroad, there are still some challenges and problems. In the future, it is necessary to further strengthen technical research and application practice, promote the widespread application of drones and artificial intelligence technology in construction management, and contribute to the digital transformation and intelligent development of the construction industry. Summary of the invention
[0005] Problems with existing technologies:
[0006] The application of drone and artificial intelligence technology is mainly concentrated in a single technical field, such as image recognition and data analysis. It is rarely used in actual construction management, and further technical research and practice are needed.
[0007] The technical solution adopted by the present invention is as follows: a visual dynamic monitoring and early warning system for bridge construction progress based on drone intelligent identification and BIM, including a model generation and viewing function module, a drone aerial photography data uploading and viewing function module, a bridge component target detection function module, and a real-time monitoring and early warning construction progress function module.
[0008] Furthermore, the model generation and viewing function module includes a BIM model generation unit based on the construction drawing of the target bridge, a 3D viewing BIM model unit, a rendering and modification BIM model unit, and a management interactive engineering BIM model unit. The BIM model generation module based on the construction drawing of the target bridge automatically constructs a BIM model skeleton by identifying and extracting features from the original construction drawing, and then adds parameterization and attribute information to the generated skeleton model, and then uses an AI optimization algorithm to optimize the generated BIM model to automatically generate a complete BIM model.
[0009] Furthermore, the human-machine aerial photography data uploading and viewing function module includes a unit for receiving aerial video and camera information, a unit for counting and viewing user uploaded data, and a unit for extracting and converting aerial camera parameters. The module for extracting and converting drone aerial video camera parameters uses AI to intelligently identify key information such as camera latitude and longitude and Euler angles in the SRT subtitle file uploaded by the user and extracts it, and then converts the data into a data structure of a JOSN object array to provide data for other front-end functions of the platform.
[0010] Furthermore, the bridge component target detection functional module includes a UI target detection platform unit, an extraction and conversion target detection and recognition result unit, and a query history detection information unit. The bridge component target detection module obtains a bridge component recognition model through training using the YOLOv5 network, performs target detection on the input bridge component aerial video, and finally identifies all bridge components and their construction status in the aerial video of the construction site. The extraction and conversion target detection and recognition result module extracts the category, confidence and relative pixel coordinates of the target detection result through a program, and provides data for the platform's real-time monitoring and early warning construction progress functional module.
[0011] Furthermore, the real-time monitoring and early warning construction progress function module includes a simulated flight and real-time display of construction progress unit, a construction organization plan viewing unit, and a real-time construction progress dynamic early warning unit. It provides an intuitive user interface through the secondary development function of the BIM model rendering platform, displays real-time construction progress, component status, and comparison between virtual models and actual aerial images, supports user interactive roaming, and allows users to understand the details of the construction site through virtual scenes. The simulated flight and real-time display of construction progress module is implemented by integrating three API functions: simulated flight function, virtual scene component matching function, and coloring function. The real-time construction progress dynamic early warning module warns of the completion status of construction by comparing the real-time displayed construction progress with the progress required by the construction organization plan.
[0012] Furthermore, the training of the bridge component recognition model includes the following steps:
[0013] Step 1: Obtain aerial videos of bridge construction sites from bridge construction units and public websites, capture a frame of the video every 3 seconds, and form a bridge component recognition model dataset;
[0014] Step 2: Perform data enhancement on the data set by adding noise, changing brightness, cropping, translation, rotation, mirroring, cutout, etc. to expand the bridge component data set;
[0015] Step 3: Use annotation tools to annotate the bridge components in the image. The annotation categories include foundation (completed), foundation (under construction), pier (completed), pier (under construction), cap beam (completed), cap beam (under construction), span structure (completed), span structure (under construction), a total of 8 categories of targets. 80% of the images are randomly selected as training sets and 20% of the images are used as test sets.
[0016] Step 4: Improve the YOLOv5l model network by adding the EMA attention mechanism in C3, replacing NMS with Soft-NMS, changing the removal of other bounding boxes in the NMS algorithm to modifying the confidence of other bounding boxes, and optimizing the loss function to MPDloU;
[0017] Step 5: Use the prepared dataset to train the YOLO model, select best.pt as the weight of the recognition model, use the test set to evaluate the performance of the model, and save the last.pt model weight to maintain the learning ability of the model, ensuring that the model can be upgraded later by adding training data.
[0018] Furthermore, the function of simulating flight and displaying the construction progress unit in real time is realized by the following steps:
[0019] Step 1: Obtain the bounding box information and center point coordinates of all components in the BIM model component dataset;
[0020] Step 2: Import the camera parameter data of the drone’s campos and camdir, and use the mobile camera perspective API function of the BIM model rendering platform to simulate roaming along the actual drone aerial photography route in the model viewing page;
[0021] Step 3: Use the BIM model rendering platform to convert world coordinates into screen position API functions to determine whether the component is in the real-time roaming perspective, and calculate the relative position and pixel area of the component in the real-time roaming perspective;
[0022] Step 4: Use the BIM model rendering platform API function to classify and color the components according to the component ID and construction status information;
[0023] Step 5: Obtain relevant information of the matched identified component by comparing the pixel area and relative position of the detection frame with the relative position and pixel area in the roaming view.
[0024] Step 6: According to the relevant information of the identified components, the component height and component position are calculated, the related components are associated, and the corresponding component information is recorded.
[0025] Step 7: Integrate functions to achieve real-time simulated flight display progress. Simulate the drone flight effect and color the model in the camera's view according to the recognition content to show the construction progress.
[0026] Furthermore, the method for visual dynamic monitoring and early warning of bridge construction progress based on UAV intelligent identification and BIM is implemented, comprising the following steps:
[0027] Step 1: Generate the BIM model of the project based on the target bridge construction drawing provided by the bridge design unit;
[0028] Step 2: Import the generated BIM model into the BIM model rendering platform, provide a web link, and create a page that can interact with the BIM model in real time on the web page;
[0029] Step 3: Upload the aerial video of the bridge construction site taken at the construction site and the camera parameter file of the video;
[0030] Step 4: Extract the camera parameters of the drone aerial video, including latitude and longitude and Euler angles, convert the latitude and longitude into the position format corresponding to the BIM model rendering platform, and convert the Euler angle into the camera orientation format;
[0031] Step 5: Use the bridge component recognition model to identify the aerial video, capture a frame every 3 seconds as a checkpoint, and obtain the recognition results of each frame: bounding box coordinates, category, and confidence level;
[0032] Step 6: Convert the pixel coordinates of the image bounding box into the relative position of the center point of the bounding box in the image, and calculate the pixel area of the detection box;
[0033] Step 7: Organize component information. Organize the identified information in the form of a JSON format object array data structure. Each component contains its category, status, confidence, and relative position in the image;
[0034] Step 8: By simulating the actual flight route of the drone and roaming in the BIM model with the same perspective, the drone video and the virtual screen in the BIM model are compared in real time. According to the information in the screen, the BIM model is updated in real time to dynamically display the BIM model corresponding to the actual construction site;
[0035] Step 9: Count the number of completed components, components under construction and unfinished components in the BIM model, and generate a construction progress task quantity progress chart;
[0036] Step 10: Import the construction organization schedule of the project to visualize the total construction volume and expected construction progress;
[0037] Step 11: Compare the task schedule with the construction organization plan, and based on the real-time completion status, promptly warn the user whether the construction progress is in line with expectations.
[0038] After adopting the above technical solution, the beneficial effects of the present invention are as follows:
[0039] 1. The present invention adopts drone aerial photography technology, artificial intelligence algorithm and BIM technology to realize real-time monitoring and intelligent management of bridge construction progress.
[0040] 2. The present invention uses a large number of aerial photos of real bridge construction sites to train the bridge component recognition model, and optimizes and improves the traditional YOLOv5 model, which greatly increases the efficiency and accuracy of bridge component target detection.
[0041] 3. The present invention compares the camera perspective of drone aerial photography with the roaming perspective in the BIM model, and uses target detection means to compare the construction status of the construction site in real time, dynamically obtains real-time progress data of the construction site, and significantly improves the efficiency and accuracy of the construction progress of the visualized bridge construction site.
[0042] 4. A smart management system has been built to provide more convenient tools and methods for construction management, and realize remote monitoring and real-time data analysis of the construction site. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] Figure 1 This is a schematic diagram of the overall structure of a bridge construction progress visualization dynamic monitoring and early warning method based on UAV intelligent identification and BIM proposed by the present invention;
[0044] Figure 2 This is a flow chart of the bridge component recognition model training of a bridge construction progress visualization dynamic monitoring and early warning system based on UAV intelligent recognition and BIM proposed by the present invention;
[0045] Figure 3 The present invention proposes a bridge construction progress visualization dynamic monitoring and early warning system based on UAV intelligent identification and BIM, which simulates flight and displays the construction progress flow chart in real time;
[0046] Figure 4This is a schematic diagram of the overall structure of a bridge construction progress visualization dynamic monitoring and early warning system based on drone intelligent identification and BIM proposed by the present invention;
[0047] Figure 5 The present invention provides a method and process flow diagram of an embodiment of a method and system for visual dynamic monitoring and early warning of bridge construction progress based on UAV intelligent identification and BIM.
[0048] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION
[0049] The technical solutions in the embodiments of the present invention are clearly and completely described below. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.
[0050] Embodiment:
[0051] Combine the following Figure 5 Taking an urban expressway overpass as an example, the method of the present invention is applied to monitor and warn the progress of bridge construction, and a visual dynamic monitoring and early warning method and system for bridge construction progress based on drone intelligent recognition and BIM are described, including the following steps:
[0052] Step 1: Since the project already has a ready-made BIM model, the step of generating a BIM model from the target bridge construction drawing can be skipped. The BIM model can be directly imported into the Black Hole Engine BIM rendering platform, and the Black Hole Engine API function can be used to embed the web page in the model generation and viewing function module of the bridge construction intelligent management system.
[0053] Step 2: Use a DJI drone to take aerial photos of the bridge construction site and obtain aerial video of the entire construction section.
[0054] Step 3: Upload the aerial video of the construction site to the UAV aerial data upload and viewing function module, extract the subtitle file of the aerial video, import the subtitle file into the UAV aerial video camera parameter extraction and conversion module, and obtain the UAV's latitude and longitude, height, Euler angle and other camera parameters.
[0055] Step 4: Perform target detection on the uploaded construction site video in the bridge component target detection function module, capture a frame every 3 seconds as a checkpoint, and obtain the recognition results of each frame: bounding box coordinates, category, and confidence.
[0056] Step 5: Convert the pixel coordinates of the image bounding box into the relative position of the center point of the bounding box in the image and calculate the pixel area of the detection box.
[0057] Step 6: Organize the identified information in the form of a JSON format object array data structure. Each component contains its category, status, confidence, and relative position in the image.
[0058] Step 7: By simulating the actual flight route of the drone and roaming in the BIM model with the same perspective, the pixel area and relative position of all components in the roaming perspective are obtained, and the pixel area and relative position identified by the target detection are compared, and the BIM model is updated in real time according to the information in the picture. Completed components are painted green, components under construction are painted yellow, and unfinished components are painted gray with a transparency of 50%. For components that cannot be identified due to occlusion, the construction status of the components is judged according to the logical relationship during the construction of the components, and the model is supplemented with color again, so as to dynamically display the BIM model corresponding to the actual construction site.
[0059] Step 8: Count the number of completed components, components under construction and unfinished components in the BIM model, and generate a construction progress task quantity progress chart.
[0060] Step 9: Import the construction organization schedule of the project to visualize the total construction workload and expected construction progress.
[0061] Step 10: Compare the task schedule with the construction organization schedule. Quantify the progress of the project and issue warnings for tasks that are not completed in time.
[0062] The model generation and viewing function module is used to generate and view the BIM model based on the target bridge construction drawing.
[0063] The UAV aerial photography data upload and viewing function module is used to store and view the UAV aerial photography videos and camera parameters of the project.
[0064] The bridge component target detection function module is used to perform target detection of bridge components in aerial videos of construction sites.
[0065] The real-time monitoring and early warning construction progress function module is used to carry out real-time monitoring and early warning of the current bridge construction progress.
[0066] The bridge component target detection function module includes:
[0067] The UI target detection platform unit is used to perform frame-by-frame target detection on aerial videos of bridge construction sites.
[0068] The target detection and recognition result extraction and conversion unit is used to convert the target detection results into JSON object array data format and save them.
[0069] The query history detection information unit is used to store the previous target detection results, which can be viewed and called at any time.
[0070] The real-time monitoring and early warning construction progress function module includes:
[0071] The simulated flight and real-time display of construction progress unit is used to simulate the actual flight route of the drone and roam in the BIM model with the same perspective, showing the real-time construction progress, component status and comparison between the virtual model and the actual aerial image.
[0072] The construction organization plan viewing unit is used to visually view the construction organization plan of the current project.
[0073] The real-time construction progress dynamic early warning unit is used to compare the construction progress with the construction organization plan and quantitatively analyze the progress of the project.
[0074] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.
[0075] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
[0076] The present invention and its implementation methods are described above, and such description is not restrictive, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design structures and embodiments similar to the technical solution without creativity, which should all fall within the protection scope of the present invention.
Claims
1. A visual dynamic monitoring and early warning system for bridge construction progress based on drone intelligent identification and BIM, Features: It includes model generation and viewing function modules, drone aerial photography data upload and viewing function modules, bridge component target detection function modules, and real-time monitoring and early warning construction progress function modules; The model generation and viewing function module includes a BIM model generation unit based on the construction drawing of the target bridge, a 3D viewing BIM model unit, a rendering and modification BIM model unit, and a management interactive engineering BIM model unit; The human-machine aerial photography data uploading and viewing function module includes a unit for receiving aerial photography videos and camera information, a unit for counting and viewing user uploaded data, and a unit for extracting and converting aerial photography camera parameters; The bridge component target detection function module includes a UI target detection platform unit, an extraction and conversion target detection recognition result unit, and a historical detection information query unit. The bridge component target detection module uses a bridge component recognition model obtained by training with a YOLOv5 network to perform target detection on the input bridge component aerial video; The real-time monitoring and early warning construction progress function module includes a flight simulation and real-time display of construction progress unit, a construction organization plan viewing unit, and a real-time construction progress dynamic early warning unit.
2. According to claim 1, a visual dynamic monitoring and early warning system for bridge construction progress based on drone intelligent identification and BIM, Features: The training of the bridge component recognition model comprises the following steps: Step 1: Obtain aerial videos of bridge construction sites from bridge construction units and public websites, capture a frame of the video every 3 seconds, and form a bridge component recognition model dataset; Step 2: Perform data enhancement on the data set by adding noise, changing brightness, cropping, translation, rotation, mirroring, cutout, etc. to expand the bridge component data set; Step 3: Use annotation tools to annotate the bridge components in the images. The annotation categories include 8 types of targets. 80% of the images are randomly selected as training sets and 20% of the images are randomly selected as test sets. Step 4: Improve the YOLOv5l model network by adding the EMA attention mechanism in C3, replacing NMS with Soft-NMS, changing the removal of other bounding boxes in the NMS algorithm to modifying the confidence of other bounding boxes, and optimizing the loss function to MPDloU; Step 5: Use the prepared dataset to train the YOLO model, select best.pt as the weight of the recognition model, use the test set to evaluate the performance of the model, and save the last.pt model weight to maintain the learning ability of the model, ensuring that the model can be upgraded later by adding training data.
3. According to claim 1, a visual dynamic monitoring and early warning system for bridge construction progress based on drone intelligent identification and BIM, Features: The function realization of the flight simulation and real-time display construction progress unit includes the following steps: Step 1: Obtain the bounding box information and center point coordinates of all components in the BIM model component dataset; Step 2: Import the camera parameter data of the drone’s campos and camdir, and use the BIM model rendering platform’s mobile camera perspective API function to simulate roaming along the actual drone’s aerial photography route in the model viewing page; Step 3: Use the BIM model rendering platform to convert world coordinates into screen position API functions to determine whether the component is in the real-time roaming perspective, and calculate the relative position and pixel area of the component in the real-time roaming perspective; Step 4: Use the BIM model rendering platform API function to classify and color the components according to the component ID and construction status information; Step 5: Obtain relevant information of the matched identified component by comparing the pixel area and relative position of the detection frame with the relative position and pixel area in the roaming view. Step 6: According to the relevant information of the identified components, the component height and component position are calculated, the related components are associated, and the corresponding component information is recorded. Step 7: Integrate functions to achieve real-time simulated flight display progress. Simulate the drone flight effect and color the model in the camera's view according to the recognition content to show the construction progress.
4. A visual dynamic monitoring and early warning method for bridge construction progress based on drone intelligent identification and BIM, Features: The following steps are involved: Step 1: Generate the BIM model of the project based on the target bridge construction drawing provided by the bridge design unit; Step 2: Import the generated BIM model into the BIM model rendering platform, provide a web link, and create a page that can interact with the BIM model in real time on the web page; Step 3: Upload the aerial video of the bridge construction site taken at the construction site and the camera parameter file of the video; Step 4: Extract the camera parameters of the drone aerial video, including latitude and longitude and Euler angles, convert the latitude and longitude into the position format corresponding to the BIM model rendering platform, and convert the Euler angle into the camera orientation format; Step 5: Use the bridge component recognition model to identify the aerial video, capture a frame every 3 seconds as a checkpoint, and obtain the recognition results of each frame: bounding box coordinates, category, and confidence level; Step 6: Convert the pixel coordinates of the image bounding box into the relative position of the center point of the bounding box in the image, and calculate the pixel area of the detection box; Step 7: Organize component information. Organize the identified information in the form of a JSON format object array data structure. Each component contains its category, status, confidence, and relative position in the image; Step 8: By simulating the actual flight route of the drone and roaming in the BIM model with the same perspective, the drone video and the virtual screen in the BIM model are compared in real time. According to the information in the screen, the BIM model is updated in real time to dynamically display the BIM model corresponding to the actual construction site; Step 9: Count the number of completed components, components under construction and unfinished components in the BIM model, and generate a construction progress task quantity progress chart; Step 10: Import the construction organization schedule of the project to visualize the total construction volume and expected construction progress; Step 11: Compare the task schedule with the construction organization plan, and based on the real-time completion status, promptly warn the user whether the construction progress is in line with expectations.
Citation Information
Patent Citations
BIM-based unmanned aerial vehicle monitoring method and system
CN105334862A
Construction progress visualization method based on BIM model and monitoring image comparison
CN113360583A
Bridge construction progress intelligent identification method based on improved YOLOV5S
CN116363586A
Insulator defect detection method and system based on improved YOLOv5
CN117218102A
Bridge construction progress identification method and device, equipment and medium
CN118262051A
Cited By
Intelligent prediction method for intelligent jacking construction progress of bent cap
CN121094243A
A method for intelligently predicting the construction progress of a bent cap
CN121094243B
Construction progress dynamic tracking method based on unmanned aerial vehicle
CN121279932A
