CN120278599A

CN120278599A - Intelligent 3D imaging safety quality monitoring system for building

Info

Publication number: CN120278599A
Application number: CN202510425608.7A
Authority: CN
Inventors: 罗坤
Original assignee: Individual
Current assignee: Individual
Priority date: 2025-04-07
Filing date: 2025-04-07
Publication date: 2025-07-08

Abstract

The invention relates to the technical field of building construction quality and safety management, and discloses a building intelligent 3D imaging safety quality monitoring system, which comprises an intelligent 3D imager. The intelligent 3D imaging system generates a high-precision 3D model and compares BIM model detection deviation through the cooperative work of multiple components such as an integrated intelligent 3D imaging device, an intelligent safety helmet and a WIFI coverage positioning system, the intelligent 3D imaging device acquires textures, monitors biological characteristics and interacts, the WIFI system provides a network and positions and also manages robots, an electronic datum point and a target ensure accurate modeling, a cloud control center optimizes the model, analyzes videos and generates a report, the system improves construction quality detection precision through the cooperation of the multiple components, realizes real-time synchronization and visualization of data, strengthens security management, strengthens material statistics management, solves the problem of equipment scanning blind areas and cooperation, ensures user privacy, provides convenient management, meets the requirements of intelligent, high-precision and dynamic management of the whole flow of building construction, and has remarkable advantages.

Description

Intelligent 3D imaging safety quality monitoring system for building

Technical Field

The invention relates to the technical field of building construction quality and safety management, in particular to an intelligent 3D imaging safety quality monitoring system for a building.

Background

Traditional building safety quality detection relies on manual inspection, and has the following defects:

The human error is large, the acceptance result is influenced by the technical level of personnel, experience and subjective judgment, and the data reliability is low;

The real-time performance is insufficient, the on-site data cannot be synchronized in real time, and the checking and accepting result is difficult to graphically display and trace;

The security management is weak, personnel authority is disordered in a complex environment, security accidents are frequent, and unsafe behaviors cannot be effectively identified;

The blind area and the coordination problem are that the traditional equipment has a scanning blind area, the protocols of equipment of multiple manufacturers are not uniform, and overall management is difficult.

The prior art can not meet the requirements of intelligent, high-precision and dynamic management and control of the whole process of building construction. Therefore, a need exists for a building intelligent 3D imaging security quality monitoring system that addresses the above-described limitations.

Disclosure of Invention

In order to overcome the defects in the prior art, the invention provides a building intelligent 3D imaging safety quality monitoring system, which solves the problems in the background art.

The invention provides a building intelligent 3D imaging safety quality monitoring system, which comprises the following technical scheme:

the intelligent 3D imaging instrument integrates a camera, an environment sensor, an accelerometer, a gyroscope and a laser range finder radar, generates a 3D model with millimeter-level precision based on a triangulation principle, and dynamically compares the 3D model with a preset BIM model to detect construction deviation;

the intelligent safety helmet is provided with a camera, a biosensor, a Bluetooth module and a voice interaction module, and is used for collecting construction surface texture mapping, monitoring biological characteristics of operators and identity binding, and interacting with a cloud control center in real time through wireless communication, and issuing construction notices, work forbidden areas and emergency situations through the intelligent helmet every day;

The WIFI coverage positioning system provides global network coverage and centimeter-level positioning functions, supports protocol conversion of equipment of multiple manufacturers, plans a robot operation forbidden zone, and displays the position of the intelligent safety helmet in real time in a model space;

The electronic datum points and the electronic targets provide absolute coordinate references for the 3D model, so that modeling space consistency is ensured;

When the intelligent helmet wearer mistakenly enters the forbidden area of the electronic fence, triggering warning information and dissuading the intelligent helmet wearer from the forbidden area, and guiding the intelligent helmet wearer to bypass the forbidden area;

The vehicle identification card is a data card combining Bluetooth and WiFi, and is used for communicating interaction data with WiFi through the Bluetooth, recording vehicle users, authorities, recording vehicle attributions, material types, weights, vehicle-mounted material detection reports, docking departments and contact phones.

And the cloud control center optimizes the details of the 3D model by adopting a deep learning algorithm, dynamically analyzes the video data to identify unsafe behaviors, and generates a 3D time axis model and a multidimensional management report of the whole construction process.

Further, the intelligent 3D imager comprises a fixed type imager and a handheld type imager, the fixed type imager is arranged in a construction key area, the handheld type imager is used for mobile scanning and blind area complementary modeling, the handheld type imager can also be used for displaying 3D model scenes corresponding to sites on a screen and checking the sizes, specifications, quantity and simulated live-action effects of building components, a 3D model is generated in a site material stacking area, for example, template stacking quantity is calculated, template area is calculated, the volume number of stacked steel pipes is calculated, the weight is reduced, and the like.

Further, the biosensor of the intelligent safety helmet comprises a heart rate monitoring module, and when detecting that the heart rate of the binding user is abnormal or is not matched with the identity of the binding user, the biosensor automatically triggers voice warning and pushes early warning information to a cloud control center.

Furthermore, the intelligent safety helmet is provided with an emergency help-seeking button, when a binding user encounters an emergency and needs rescue, the emergency help-seeking button is triggered manually, the intelligent safety helmet triggers voice warning and pushes warning information to a cloud control center, and the position is highlighted in the model. The intelligent safety helmet simultaneously transmits the position coordinates outwards.

Furthermore, the WIFI coverage positioning system is communicated with the third-party robot device through the multi-protocol compatible interface, the robot path is planned in real time, unauthorized devices are prevented from entering the operation restricted area, the WIFI coverage positioning system monitors and records the intelligent safety helmet moving track and the intelligent safety helmet moving position in the area, and the evacuation route is intelligently planned according to the position and guided to the safety area in emergency.

Furthermore, the cloud control center comprises a dynamic blind area complement module, blind area data are collected through an intelligent safety helmet camera, blind area characteristics are predicted by combining a deep learning algorithm, and a complete 3D model is generated.

Furthermore, the system also comprises a permission management module, wherein the permission management module is used for realizing double permission verification through binding face recognition with the safety helmet and binding mobile phone Bluetooth with the safety helmet, and triggering voice warning and security linkage response when unauthorized personnel enter a monitoring area.

Further, the cloud control center analyzes the video data in real time, extracts the motion trail, the gesture and the object state characteristics of the personnel, recognizes unsafe behaviors through a preset algorithm and pushes early warning to the intelligent safety helmet.

Further, the cloud control center analyzes the video data in real time, and distinguishes smog, abnormal temperature rise, out-of-tolerance horizontal displacement, out-of-tolerance verticality and out-of-tolerance deformation in a monitoring area, and pushes different levels of warnings according to development trends.

Furthermore, the vehicle identification card is used for controlling vehicles in the project, recording the entry time, the departure time, the material type, the weight, the vehicle-mounted material detection report, dredging the engineering vehicles and interactively recording the vehicle track of the engineering transportation materials.

Furthermore, the system supports generating a 3D completion model of the whole construction process according to a time axis, and provides a visual data basis for engineering payment, audit and quality tracing.

Furthermore, the intelligent safety helmet is internally provided with a privacy control module, so that a user is allowed to manually close the camera function, and attendance data of operators and remote assistance requests are recorded.

Furthermore, the cloud control center can call and open all cameras in the monitoring area through the advanced authority, and the cameras are arranged in the intelligent safety helmet and are used for coping with emergencies and emergency events. And a 3D live-action safe evacuation path can be generated according to the collected video and the 3D model, so that workers at all positions are guided to evacuate safely.

Furthermore, the cloud control center generates a multidimensional management report, including construction progress, quality qualification rate, attendance rate and safety event statistics, and supports real-time viewing and export of the mobile terminal, and the system can dynamically display and search the position of the intelligent safety helmet in real time in a model space.

Further, after the intelligent 3D imaging safety quality monitoring system model is built and delivered, the model is simplified to strengthen the logic relationship and the stress relationship of the components, equipment and the circuit pipelines, and the 3D model with clear delivery logic relationship and clear stress system is used for operation and maintenance reference inquiry.

The invention has the technical effects and advantages that:

the intelligent 3D imaging safety quality monitoring system for the building effectively solves the problems of traditional building safety quality detection, integrates multiple components such as an intelligent 3D imaging instrument, an intelligent safety helmet and a WIFI coverage positioning system, works cooperatively, generates a high-precision 3D model and compares BIM model detection deviation, acquires textures, monitors biological characteristics and interacts, provides a network and positioning, manages robots, ensures accurate modeling of electronic datum points and targets, spatially distinguishes forbidden areas of the electronic fences, optimizes the model, analyzes videos and generates reports by a cloud control center, improves construction quality detection precision by matching the multiple components, realizes real-time synchronization and visualization of data, strengthens security management, solves the problem of equipment scanning blind areas and cooperation, ensures user privacy, provides convenient management, and meets the requirements of intelligent, high-precision and dynamic management and control of the whole flow of the building construction.

Specifically

Improving the detection precision of construction quality

According to the invention, by means of an intelligent 3D imager, a camera, an environment sensor, an accelerometer, a gyroscope and a laser range radar are integrated, a 3D model with millimeter-level precision is generated based on a triangulation principle, and dynamic comparison is carried out with a preset BIM model. The construction deviation can be accurately detected in the process, and compared with the traditional mode of relying on manual inspection, the influence of human errors is greatly reduced, so that the construction quality detection data are more reliable. For example, in the construction of a main structure of a building, dimensional deviation of beams and columns, perpendicularity deviation of walls and the like can be accurately identified, and the construction is ensured to be strictly carried out according to design requirements, so that the overall quality and safety of the building are improved.

Enhancing real-time and data visualization

The intelligent 3D imaging instrument, the intelligent safety helmet and other devices interact with the cloud control center in real time through the WIFI coverage positioning system, and field data can be synchronized to the cloud in time. The cloud control center can not only optimize the details of the 3D model by adopting a deep learning algorithm, but also generate a 3D time axis model and a multidimensional management report of the whole construction process. Constructors and management staff can check information such as construction progress, quality conditions and the like at any time, and real-time and visual display of data is achieved.

For example, a manager can check a 3D model of the engineering progress in real time through a mobile terminal, intuitively know construction conditions of different stages, and is convenient for making decisions in time. The 3D time axis model can also provide visual data basis for engineering payment, audit and quality tracing, and is convenient for subsequent work development.

For example, a manager can also seamlessly fuse the cloud 3D model and structural construction information with a scene through mobile terminal APP positioning and camera matching, and dynamically display effects and check the sizes, specifications, quantity and positions of components without searching data from complicated drawing data. Meanwhile, the completion percentage of the construction section can be calculated through the cloud, a bill of materials is required in the construction section to be completed, and the engineering material reserve is kept in an optimal state forever.

For example, various concealed pipelines are completed in complex electromechanical engineering, and the pipeline paths, purposes and concealed prospects are automatically positioned and displayed through end positioning. And the missed connection and misconnection of each construction section and each working procedure caused by the missing connection are avoided.

Binding user teams according to intelligent safety helmets, displaying different model data and different authority data in APP, wherein the electromechanical maintenance teams are used for displaying concealed temporary pipelines, switch boxes, control boxes, valves and other routes in an emphasized manner and rapidly guiding the electromechanical maintenance teams to process faults;

enhanced security management

The intelligent safety helmet is provided with the biosensor and the camera, so that the biological characteristics of operators can be monitored, and unsafe behaviors can be effectively identified by combining dynamic analysis of the cloud control center on video data. And simultaneously, the cloud control center identifies unsafe behaviors and pushes early warning to the intelligent safety helmet by analyzing the motion trail, the gesture and the object state characteristics of the personnel. In addition, the system is also provided with a permission management module, double permission verification is realized through face recognition and mobile phone Bluetooth and safety helmet binding, and voice warning and security linkage response are triggered when unauthorized personnel enter a monitoring area. The measures ensure the personnel safety of the construction site in an all-round way, reduce the occurrence of safety accidents and improve the level of security management.

The intelligent safety helmet is used for collecting the position coordinates of the helmet at any time through the WIFI coverage positioning system, interacting coordinate information in real time at the cloud control center, dynamically displaying the coordinate information in the 3D model, and knowing the working areas and the working tracks of all staff.

Solve the blind area and the cooperative problem

The intelligent 3D imaging instrument is divided into a fixed type imaging instrument and a handheld type imaging instrument, the fixed type imaging instrument is arranged in a construction key area, the handheld type imaging instrument is used for mobile scanning and blind area complementary modeling, a dynamic blind area complementary module of the cloud control center collects blind area data through an intelligent safety helmet camera, a depth learning algorithm is combined to predict blind area characteristics, a complete 3D model is generated, and the problem of scanning blind areas existing in traditional equipment is effectively solved. Meanwhile, the WIFI coverage positioning system supports protocol conversion of equipment of multiple factories, can be communicated with third-party robot equipment, plan a robot path in real time and prohibit unauthorized equipment from entering an operation forbidden zone, so that overall management and collaborative operation of the multiple equipment are realized, and construction efficiency is improved.

Ensuring user privacy and providing convenient management

The intelligent safety helmet is internally provided with the privacy control module, so that a user is allowed to manually close the camera function, and the privacy of an operator is fully ensured. Meanwhile, the intelligent safety helmet can also record attendance data of operators and remote assistance requests, and a multidimensional management report generated by the cloud control center comprises construction progress, quality qualification rate, attendance rate, safety event statistics and the like, supports real-time checking and exporting of a mobile terminal, provides a convenient tool for construction management, and improves management efficiency.

In conclusion, the intelligent 3D imaging safety quality monitoring system for the building comprehensively solves the problems existing in the traditional safety quality detection of the building, meets the requirements of intelligent, high-precision and dynamic management and control of the whole flow of the building construction, and has remarkable technical effects and practical values.

Drawings

FIG. 1 is a diagram of a system architecture in accordance with the present invention;

fig. 2 is a schematic diagram of an implementation example application according to the present invention:

FIG. 3 is a schematic diagram of the logic of the present invention.

Detailed Description

The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.

It will be understood that the terms "first," "second," and the like, as used herein, may be used to describe various elements, but these elements are not limited by these terms unless otherwise specified. These terms are only used to distinguish one element from another element.

In order to enable those skilled in the art to better understand the technical solutions of the present application, the technical solutions of the embodiments of the present application will be clearly and completely described below, and it is obvious that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.

Examples

System deployment and hardware installation

WIFI coverage and positioning system deployment

In the construction site, signal intensity and interference source distribution of different areas are mapped in detail by special wireless signal survey software in the early stage. And according to the mapping result, layering and partitioning are adopted to lay the WIFI base station. And arranging base stations at intervals of certain floors in a main construction area of the building, and adjusting signal coverage angles by using directional antennas to ensure the signal stability of indoor and outdoor construction areas. For outdoor open fields, such as the periphery of a large foundation pit and a material stacking area, the signal strength and the positioning accuracy are met, the distance between base stations is flexibly adjusted according to the triangular or quadrilateral grid layout, and the whole area coverage without dead angles is realized.

When selecting the positioning module, the cost, the precision and the field environment are comprehensively considered. Although the UWB technology has high precision, the UWB technology can be interfered in areas with multiple metals and strong signal reflection, and the Bluetooth beacon is combined for complementary positioning. When the UWB positioning tag is installed, the UWB positioning tag is firmly fixed at the specific positions of the intelligent equipment and the personnel safety helmet, shielding is avoided, calibration compensation is carried out, and positioning errors caused by human shielding and equipment installation position difference are reduced. The bluetooth beacon is deployed at key nodes in the building, such as a building entrance and an elevator cabin, and is used for assisting in realizing more accurate indoor positioning. Meanwhile, a redundant backup network is built, a standby base station and an emergency power supply are arranged, continuity of the network and positioning service is guaranteed, automatic switching can be achieved when a main network fails, and normal operation of the system is guaranteed.

Electronic fiducial point and target placement

Before the electronic datum point is installed, the site measurement is carried out by means of a high-precision total station, and the measurement precision is accurate to the sub-millimeter level. And (3) repeatedly measuring and confirming each datum point according to a coordinate system of the building design drawing, and ensuring that the absolute coordinate accuracy error is within +/-0.5 mm. During installation, a special embedded part is adopted, and the electronic datum point is firmly embedded into a concrete foundation or a stable building structure, so that the electronic datum point is ensured not to be displaced or loosened in the construction process. After the installation is completed, the datum point is protected by using the protective shell, so that construction collision and severe weather erosion damage are prevented.

And the electronic target is subjected to customized design according to the characteristics of the construction area and the 3D modeling requirement. In the construction area of complex structures, such as large bridge piers and special-shaped building components, targets with unique optical marks and electronic signal reflecting devices are designed, so that the recognition effect of the intelligent 3D imaging instrument is enhanced. The target installation position is accurately calculated, so that the imaging devices with different angles can be effectively identified. Meanwhile, the electronic datum points and the targets are calibrated and checked regularly, the coordinates are retested by using professional measuring equipment, the accuracy of the coordinates is ensured to always meet the requirements, and reliable coordinate datum points are provided for 3D modeling.

Intelligent 3D imager installation

When the fixed imager is installed at the tops of the tower crane and the scaffold, the special damping installation support is adopted, so that the influence of equipment vibration on the imaging precision is reduced. When the tower crane is installed, the installation angle and the installation position are adjusted, so that the field of view of the tower crane covers the whole building construction surface, and the rotating and lifting functions of the tower crane are combined, so that the dynamic scanning in a wider range is realized. When the scaffold is installed at the top, the scaffold is constructed and removed, and the fixing device which can be rapidly disassembled and reinstalled and can ensure the installation accuracy is designed.

The handheld imager is personalized before being equipped to quality inspectors. And adjusting the layout of the operation buttons and the menu display mode of the equipment according to the use habit of the quality inspector. Meanwhile, the portable charging equipment and the data storage module are provided for the portable charging equipment, so that the portable charging equipment is convenient to use for a long time on a construction site. Aiming at different construction scenes, such as narrow space and high-altitude operation, the auxiliary tool matched with the handheld imager is customized, such as an extension rod and a wide-angle lens accessory, so that image data of a special position can be conveniently acquired.

And calculating materials by using 3D imaging, for example, calculating the stacking amount of the templates and calculating the area of the templates. And calculating the volume number of the pile steel pipe roots and the weight. The layout of the site is reasonably planned, and the layout of the line water pipes is planned.

Intelligent safety helmet configuration

Before distribution, the intelligent safety helmet carries out comprehensive performance detection on components such as a built-in camera, a heart rate sensor, a voice module, a Bluetooth module, an emergency help seeking module and the like. And (3) calibrating the heart rate sensor by using professional sensor calibration equipment, so as to ensure that the measurement accuracy error is within +/-2 times/min.

The intelligent safety helmet can simultaneously start intercom and group chat functions and call video of the opposite side cameras, so that barrier-free communication among floors is realized under the noise of building equipment, and safety accidents caused by unsmooth communication and incorrect operation are avoided. Such as multi-person collaborative cable laying, equipment command lifting and the like

The security of identity verification is ensured through face recognition and mobile phone Bluetooth binding right. Face recognition equipment is arranged at the entrance of a construction site, face and intelligent safety helmet IP binding is arranged, safety helmet can be retrieved through a mobile phone, voice warning and security linkage response are triggered when an unauthorized safety helmet or an IP-mismatched intelligent safety helmet is worn in a construction area, and identity recognition and authority verification are automatically completed when workers enter, so that unauthorized personnel cannot enter the safety helmet. Multiple encryption mechanisms are arranged for the intelligent safety helmet, and the collected data are transmitted and stored in an end-to-end encryption mode, so that the safety of the data is guaranteed, and privacy leakage is prevented. Meanwhile, the software system of the helmet is updated regularly, the performance and the functions of the equipment are improved, and new safety monitoring and interaction functions are added.

Data acquisition and model construction

Surface texture map acquisition

When constructors wear intelligent helmets to collect surface texture data, the system automatically adjusts the collection strategy according to construction progress and regional characteristics. In the construction stage of the main structure of the building, the integral textures of the wall surface and the beam column are collected in a focus mode, and in the decoration stage, the surface textures of the decorative materials are collected in a fine mode. In order to improve the collection efficiency and quality, an intelligent collection guide program is developed, and constructors are guided to collect at proper distances and angles through a built-in display screen and voice prompts of the helmet.

In the acquisition process, an image enhancement algorithm is utilized to process the acquired image in real time, so that the definition and contrast of the image are improved. Meanwhile, images acquired at different angles are spliced into a complete texture map in a seamless mode by adopting an image splicing technology, and continuity and accuracy of textures are guaranteed. The collected data is uploaded to the field recording control center in real time through WIFI, and in the uploading process, a data compression technology is adopted, so that the data transmission quantity is reduced, the transmission speed is improved, and meanwhile, the data quality is guaranteed not to be lost.

3D scanning and preliminary modeling

The fixed 3D scanning equipment is arranged below a tower crane arm, scanning from all angles is continuously repeated by utilizing tower crane swing, 3D model data is continuously corrected and perfected through an AI algorithm, time axis information is added according to an administrator set time interval, and in the scanning process, clear and accurate scanning data can be obtained in the daytime and the night by combining a camera, an environment sensor, an accelerometer, a gyroscope and a laser ranging radar through AI algorithm calculation optimization.

When the handheld imager performs indoor or local fine scanning, the position information of the current scanning area in the whole model is acquired through data interaction with the fixed imager, so that seamless fusion of data is realized. Aiming at complex structures such as reinforced concrete nodes and pipeline dense areas, a multi-view scanning method is adopted to collect data from different angles, and a point cloud fusion algorithm is utilized to merge point cloud data of multiple views to construct a more complete and accurate local model.

Cloud model optimization

When the on-site recording control center uploads data to the cloud control center, an asynchronous transmission and breakpoint continuous transmission technology is adopted, so that stability and integrity of data transmission are ensured. And after the cloud control center receives the data, optimizing the 3D model by using a deep learning algorithm. In the optimization process, a countermeasure network (GAN) generation technology is introduced to generate more real and fine model details, so that the visualization effect of the model is improved.

When the BIM model is compared with the BIM model, not only the geometry, the size, the position and the number of the components are compared, but also the material property and the spatial position relation of the components are comprehensively compared. For deviation data, hierarchical classification management is adopted, and slight deviation, general deviation and serious deviation are classified according to the size and type of deviation and the influence degree on engineering quality. Different early warning is triggered according to the deviation level and the emergency degree, reports with different detailed degrees are generated aiming at the deviation of different levels, and accurate basis is provided for subsequent rectifying work.

Real-time monitoring and safety control

Behavior recognition and early warning

When the security cameras and the intelligent helm monitor cooperatively, the images of cameras at different positions are fused by utilizing a multi-camera fusion technology, so that the monitoring range is enlarged, and the monitoring blind area is reduced. In the behavior recognition algorithm, a technology based on a space-time convolution network (STCN) is adopted, so that not only can the static gesture of the personnel be recognized, but also the motion trail and the motion sequence of the personnel can be analyzed, and the accuracy and the instantaneity of behavior recognition are improved.

The security camera and the intelligent helmet cooperatively monitor, the face and the intelligent helmet are bound together through IP, and voice warning and security linkage response are triggered when the authorized helmet is not worn in a construction area and a monitoring range and the helmet is not matched with the face.

When unsafe behavior or emergency is detected, the cloud control center sends voice warning to the corresponding helmet, and meanwhile the construction staff is attracted through various modes such as vibration reminding, lamplight flickering and the like. When the management terminal flicks the screen for early warning, detailed early warning information including the time, place, personnel information and behavior type of unsafe behavior is displayed, and processing suggestions, relevant safety standard links and intelligent guiding evacuation routes are provided, so that the management personnel can process the information in time. Meanwhile, the unsafe behavior events are recorded into a safety management database, statistical analysis is carried out, the high-incidence area and the time period of the potential safety hazard are found, and data support is provided for formulating targeted safety management measures.

Biological feature monitoring

The heart rate sensor built in the intelligent helmet adopts a photoelectric volume pulse wave (PPG) technology, and combines an advanced filtering algorithm to remove noise generated by movement, electromagnetic interference and the like, so that the physiological state of a worker is accurately measured. When an abnormality in heart rate is detected, the system determines the type of abnormality based on a preset heart rate variation curve and a threshold, e.g., too fast heart rate may be due to high intensity of labor, emotional stress, or physical illness, and too slow heart rate may be physical fatigue or cardiovascular problems.

The system takes different countermeasures according to different anomaly types. For mild anomalies, such as a brief increase in heart rate, the helmet sends out a mild voice prompt to advise the worker to rest properly, and for severe anomalies, such as a sustained exceeding of the risk threshold, the system immediately notifies the site safety personnel and medical emergency personnel, and provides real-time location and health information of the worker to ensure timely rescue. At the same time, the biological characteristic data of workers are recorded into the health management file, and references are provided for long-term health assessment and labor arrangement.

Robot collaboration management

When communication is established with the third party robot, a universal communication protocol conversion middleware is developed, and the communication protocol conversion middleware is compatible with communication protocols of various robots. Through middleware, remote control and state monitoring of the robot are realized, and the working state, position information and task progress of the robot are obtained in real time. When the robot operation forbidden zone is defined, a virtual boundary of the forbidden zone is constructed by utilizing a 3D modeling technology and is associated with the intelligent 3D imager and a positioning system of the intelligent safety helmet.

When the intelligent helmet wearer mistakenly enters the electronic fence forbidden zone, the system firstly alerts personnel to leave through helmet voice warning and vibration, and simultaneously sends a stop instruction to the robot. If the personnel do not respond in time, the system automatically controls the robot to adjust the working mode, such as reducing the working speed and suspending dangerous actions, so as to avoid collision accidents. In addition, in the operation process of the robot, the distance between the robot and surrounding personnel and equipment is monitored in real time, and when the distance is too close, an early warning signal is sent out, so that the construction safety is ensured.

Blind area dynamic complement and model calibration

Blind zone identification and completion

The cloud control center utilizes a machine learning algorithm to carry out depth analysis on historical scanning data, not only considers the position, the visual angle and the scanning range of the fixed imager, but also predicts the position of a blind area by combining environmental factors of a construction site, such as equipment layout and personnel activity. By adopting a target detection algorithm based on deep learning, objects which possibly cause shielding, such as large-scale construction equipment, building components and the like, are identified, and the blind area range is more accurately determined.

When the intelligent helmet nearby is instructed to start a camera to supplement and collect data, the most suitable helmet is selected to execute the task according to the position of the helmet, the electric quantity, the task priority and other factors through an intelligent task allocation algorithm. In the acquisition process, the quality of the supplementary data is improved by utilizing the image enhancement and super-resolution reconstruction technology. And after the acquisition is completed, adopting a data fusion algorithm to fuse the supplementary data with the original 3D model, and ensuring the integrity and accuracy of the model.

Model consistency check

After the construction is finished every day, when the system compares the 3D model with the BIM design drawing, a comparison method based on feature point matching and semantic segmentation is adopted. Firstly, extracting key feature points in the model, such as edges, corner points and the like of the components, carrying out accurate matching, and then classifying and identifying different components in the model by utilizing a semantic segmentation technology, and comparing the attributes and the position relations of the components.

When the deviation report is generated, not only the numerical value of the deviation is listed, but also the position and the shape of the deviation in the model are displayed in a visual mode, so that the manager can intuitively understand the deviation. For areas with larger deviation, detailed correction suggestions and simulation correction effects are provided, and constructors are helped to formulate correction schemes. Meanwhile, deviation data are fed back to a quality control system, construction technology and construction flow are optimized, and similar deviation is avoided.

Data management and application

3D timeline model generation

When the cloud end updates the 3D model every day according to the construction progress, a version control technology is adopted, and the time, the updating content and the updating personnel information of each model update are recorded. Through the time axis model, not only can the 3D model state of each stage in the construction process be checked, but also the comparison analysis of different versions of models can be carried out, and the engineering change condition can be known.

In order to facilitate the viewing and operation of users, a 3D model visual interface based on the WebGL technology is developed, the 3D time axis model is supported to be directly browsed in a browser, and no additional software is required to be installed. In the visual interface, various interactive functions such as model scaling, rotation and translation, component information inquiry, history version switching and the like are provided, so that the requirements of different users are met.

Multidimensional report output

And when the qualification rate and the deviation correction completion rate of the component are counted, the quality report forms adopt a data mining technology to carry out deep analysis on the quality data. Not only the overall qualification rate and the correction completion rate are counted, but also classification statistics is carried out according to the multidimensional degree such as the construction area, the construction team, the component type and the like, and the distribution rule of the quality problem is found out. And (3) carrying out key analysis on the areas or groups with large quality fluctuation, finding out reasons and making improvement measures.

When the type, the frequency and the processing result of the early warning event are recorded in the safety report, the association analysis algorithm is utilized to find out the association relation between different safety events, such as the association between certain unsafe behavior and specific construction environment and equipment faults. By analyzing the association relations, the safety risk is predicted in advance, and preventive measures are taken to reduce the probability of occurrence of safety accidents.

When the attendance report automatically generates personnel attendance records based on the helmet positioning data, the attendance data is intelligently analyzed by combining the construction task arrangement and the working time requirements. Not only the attendance time of the personnel is recorded, but also the working track and the residence time of the personnel on the construction site are analyzed to judge whether the conditions of passive work failure, post-shifting and the like exist. Meanwhile, when the attendance data are in butt joint with the wage system, automatic accounting and distribution of wages are realized according to different attendance conditions and work performances.

Mobile end interaction

When a manager views the real-time model through the mobile phone client, the self-adaptive loading technology is adopted, and the loading precision and the display effect of the model are automatically adjusted according to the network condition and the equipment performance of the mobile phone. When the early warning information is received, the early warning information is timely pushed to a manager through a message pushing mechanism of the mobile phone, so that the information is ensured not to be missed. Meanwhile, the mobile phone client is provided with the warning information classification reminding function, and a manager can set different reminding modes and priorities for different types of warning information according to own requirements.

When the remote control helmet camera is used for video inspection or assisting workers in solving the technical problem, a real-time video streaming transmission technology is adopted, so that the fluency and instantaneity of video pictures are ensured. In the video call process, the voice enhancement and noise reduction technology is utilized to improve the call quality. Meanwhile, a gesture operation function of the mobile phone client is developed, and a manager can control the helmet camera through simple gesture operation, such as adjusting a shooting angle, enlarging and reducing pictures and the like, so that the convenience of operation is improved. The virtual queuing system of the reservation equipment can be realized by utilizing mobile terminal interaction, the virtual queuing system queues according to a progress plan, a priority level of a team and an emergency degree set by an administrator, the team can check own queuing progress and equipment construction position in real time, time is reasonably arranged, and when the equipment is idle, the system can broadcast by the whole person, so that the equipment is more efficient and reasonable to use. If the fight event of the team robbery tower crane is avoided frequently in the construction site.

The mobile phone client can utilize the space coordinates of the App and the intelligent helmet to scan the area pre-display space structure or reproduce the concealed equipment, components, pipelines and the like through the mobile phone camera. The life cycle of the system is the whole life cycle from building to operation to end.

System maintenance and extension

Protocol compatibility upgrade

When the cloud reserves an open interface, international standards and industry specifications, such as the internet of things communication protocols MQTT, coAP and the like, are followed, so that the universality and compatibility of the interface are ensured. And the development dynamic of the third party equipment is regularly concerned, and the protocol conversion module is updated in time to support the access of the newly-appearing equipment. Before the new equipment is accessed, strict compatibility tests are carried out, including communication stability tests, data interaction tests, function cooperative tests and the like, so that the normal operation of the system is not affected after the new equipment is accessed.

And establishing a device access management platform to perform centralized management on the third-party devices of all access systems. On the platform, the basic information, access time, use state, maintenance record and the like of the equipment are recorded, so that the equipment is convenient to monitor and manage. Meanwhile, a device authority management mechanism is set, different authorities are distributed to the device according to the type, the function and the use scene of the device, and the safety of the system is ensured.

Algorithm iteration

When the deep learning model is continuously optimized based on historical data, an online learning and incremental learning technology is adopted, so that the model can learn new data in real time, and the performance is continuously improved. New data of the construction site including safety event data, quality deviation data, personnel behavior data and the like are collected regularly, and the model is trained and optimized. Meanwhile, the expert in the field is invited to evaluate and guide the model, and the structure and parameters of the model are adjusted according to the opinion of the expert, so that the accuracy and reliability of the model are improved.

Focusing on the latest research results in the field of artificial intelligence, new algorithms and techniques are introduced into the system in time. For example, the object detection algorithm based on deep learning is upgraded from a traditional convolutional neural network to a more advanced transducer-based object detection algorithm, so that the accuracy of behavior recognition and object detection is improved. When a new algorithm is introduced, sufficient testing and verification are performed to ensure the compatibility and stability of the system with the existing system.

Example of implementation

Taking a certain high-rise building construction as an example:

In the foundation pit stage, firstly, the construction site is comprehensively surveyed and planned, and the installation position of the electronic datum point is determined. The key position of the foundation pit is measured and marked by using a high-precision measuring instrument, then an electronic datum point is installed, and the coordinate precision of the electronic datum point is ensured by laser calibration.

The fixed imager is arranged at the high points around the tower crane and the foundation pit, scans the foundation structure and generates an initial 3D model. In the scanning process, parameters of the imager are adjusted and optimized, so that high-quality and high-accuracy scanning data are obtained.

During decoration construction, workers wear the helmet to collect wall textures. The hand-held imager scans the rebar junction. When the constructor collects data, the constructor is strictly carried out according to the operation standard, so that the collected data is accurate and complete.

Cloud daily compares the model with BIM design, and timely discovers deviation in the construction process. For example, when a floor column is detected to deviate by 10mm, the system immediately pushes a correction instruction, and a manager checks the deviation position through the mobile terminal and organizes related personnel for analysis and discussion to make a correction scheme.

When the worker does not wear the safety rope, the helmet sends out warning information in real time, and the event is recorded to a safety report. And safety manager carries out safety education and training on related workers according to the safety report, so that the safety consciousness of the workers is improved.

In the decoration stage, the mobile phone client APP is utilized to check the hidden pipeline, so that the damage to the hidden line caused by the installation of subsequent components is avoided.

In the completion stage, the time axis model is used for completely recording the whole construction process, providing visual basis for government quality supervision departments and audits, and government quality supervision and auditors can intuitively know the progress situation and quality condition of construction through the time axis model to complete comprehensive and accurate quality assessment and cost audit report of the engineering.

In the operation stage, a mobile phone client APP is utilized to inquire fault cable routing, fault equipment position, power supply routing and the like

The foregoing examples illustrate only a few embodiments of the invention and are described in detail herein without thereby limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention. Accordingly, the scope of protection of the present invention is to be determined by the appended claims.

The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, and alternatives falling within the spirit and principles of the invention.

Claims

1. A building intelligent 3D imaging safety quality monitoring system, characterized by:

Intelligent 3D imager, integrating camera, environmental sensor, accelerometer, gyroscope and laser rangefinder radar, generates 3D model with millimeter-level accuracy based on triangulation principle, and dynamically compares it with the preset BIM model to detect construction deviation;

Smart safety helmets are equipped with cameras, biosensors, Bluetooth modules and voice interaction modules to collect construction surface texture maps, monitor operator biometrics, and identity binding. They also interact with the cloud control center in real time through wireless communications, and publish construction notices, work restricted areas and emergency situations through smart helmets every day.

The WIFI coverage positioning system provides full-area network coverage and centimeter-level positioning functions, supports protocol conversion of multi-manufacturer equipment, plans robot operation restricted areas, and displays the position of the smart safety helmet in real time in the model space;

Electronic reference points and electronic targets provide absolute coordinate references for 3D models to ensure consistency in modeling space;

Electronic fence: When the wearer of the smart helmet enters the restricted area of the electronic fence by mistake, a warning message will be triggered to persuade the wearer to leave, and the wearer of the smart helmet will be guided to bypass the prohibited area;

Vehicle identification card, a data card that combines Bluetooth and WiFi, exchanges data through Bluetooth and WiFi communications, records vehicle users, permissions, vehicle ownership, material types, weight, on-board material inspection reports, docking departments and contact numbers;

The cloud control center uses deep learning algorithms to optimize 3D model details, dynamically analyze video data to identify unsafe behaviors, and generate a 3D timeline model and multi-dimensional management reports for the entire construction process.

2. According to claim 1, a building intelligent 3D imaging safety and quality monitoring system is characterized in that: the intelligent 3D imager includes two types: fixed and handheld. The fixed imager is deployed in key construction areas, and the handheld imager is used for mobile scanning and blind spot supplementary modeling. The handheld imager can also present a 3D model scene corresponding to the site on the screen, which is used to check and verify the size, specification, quantity and simulate the real scene effect of the building components, and generate a 3D model in the material stacking area of the site.

3. According to claim 1, a building intelligent 3D imaging safety quality monitoring system is characterized in that: the biosensor of the smart safety helmet includes a heart rate monitoring module. When it is detected that the heart rate of the bound user is abnormal or does not match the identity of the bound user, a voice warning is automatically triggered and the warning information is pushed to the cloud control center.

4. A building intelligent 3D imaging safety quality monitoring system according to claim 1, characterized in that: the WIFI coverage positioning system communicates with third-party robot equipment through a multi-protocol compatible interface, plans the robot path in real time and prohibits unauthorized equipment from entering the operation restricted area, and the WIFI coverage positioning system monitors and records the activity trajectory and position of the smart safety helmet in the area. In an emergency, the evacuation route is intelligently planned according to the location and guides to a safe area.

5. According to claim 1, a building intelligent 3D imaging safety quality monitoring system is characterized in that: the cloud control center includes a dynamic blind spot completion module, which collects blind spot data through an intelligent safety helmet camera, predicts blind spot features in combination with a deep learning algorithm, and generates a complete 3D model.

6. According to claim 1, a building intelligent 3D imaging safety quality monitoring system is characterized in that: the system also includes a permission management module, which realizes double permission verification through face recognition and helmet binding, mobile phone Bluetooth and helmet binding, and triggers voice warning and security linkage response when unauthorized personnel enter the monitoring area.

7. A building intelligent 3D imaging safety quality monitoring system according to claim 1, characterized in that: the cloud control center analyzes video data in real time, extracts personnel movement trajectory, posture and object state characteristics, identifies unsafe behaviors through preset algorithms and pushes warnings to smart safety helmets.

8. According to claim 1, a building intelligent 3D imaging safety and quality monitoring system is characterized in that: the system supports the generation of a 3D completion model of the entire construction process according to the timeline, providing a visual data basis for project payment, auditing and quality traceability.

9. According to claim 1, a building intelligent 3D imaging safety quality monitoring system is characterized in that: the smart safety helmet has a built-in privacy control module, allowing the user to manually turn off the camera function and record the operator's attendance data and remote assistance requests.

10. According to claim 1, a building intelligent 3D imaging safety and quality monitoring system is characterized in that: the cloud control center generates multi-dimensional management reports, including construction progress, quality pass rate, attendance rate and safety event statistics, and supports real-time viewing and exporting on mobile terminals. The system can dynamically display and search for the location of smart safety helmets in real time in the model space.