Building engineering construction site robot safety inspection and early warning method based on machine vision

Abstract

The invention discloses a construction engineering construction site robot safety inspection and early warning method based on machine vision, and the method employs a tracked robot, integrates a plurality of sensors, employs an improved autonomous navigation and positioning technology, and combines a deep learning driven machine vision detection technology. And real-time monitoring and potential safety hazard identification of the construction site are realized. The system further comprises a real-time early warning system, a cloud platform data analysis and user customization interface, and a complete automatic safety monitoring solution is formed. According to the method, the inspection efficiency and accuracy are remarkably improved, the safety risk of manual inspection is reduced, and innovative technical support is provided for safety management of a building engineering construction site.

Description

Technical Field

[0001] The present invention relates to the technical field of building construction, and in particular to a machine vision-based robot safety inspection and early warning method for a construction site of a building engineering project. Background Art

[0002] With the continuous development of construction projects, construction site safety management has become increasingly important. Traditionally, construction site supervision and management relies primarily on manual labor, which has many limitations, including long workloads, low efficiency, safety hazards caused by human error, and high error rates during manual inspections, making it difficult to accurately and efficiently complete site inspections. To address these issues, intelligent inspection robots based on machine vision have emerged. Summary of the Invention

[0003] In order to solve the technical problems existing in the above-mentioned prior art, the present invention provides a robot safety inspection and early warning method for construction engineering construction sites based on machine vision, which not only improves the inspection efficiency and accuracy, but also greatly reduces the safety risks of manual inspection, providing strong technical support for the safety management of construction sites.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: a machine vision-based robot safety inspection and early warning method for construction engineering construction sites. This method uses a crawler robot integrated with multiple sensors, utilizes improved autonomous navigation and positioning technology, and combines it with deep learning-driven machine vision detection technology to achieve real-time monitoring of construction sites and identification of safety hazards. The crawler robot body integrates advanced sensor systems such as a high-resolution color camera, a precise depth camera, and an IMU.

[0005] Autonomous navigation and positioning, using lidar data to build a 3D map, and achieving precise positioning through an improved Kalman filter algorithm with adaptive parameter adjustment and SIFT feature point matching technology;

[0006] Machine vision inspection uses U-Net neural networks and YOLOv3 networks optimized for construction scenarios and trained for at least 1,000 iterations;

[0007] A real-time early warning system that integrates an audible and visual alarm system that automatically adjusts to the level of urgency and a remote video monitoring platform;

[0008] Cloud platform data analysis, with data encryption and access control functions to ensure information security;

[0009] The user interface focuses on user experience and accessibility, providing an intuitive touch operation interface and voice command input function. The interface style and functional modules are personalized according to the needs of different user groups.

[0010] The specific steps include:

[0011] 1. The crawler robot body is designed with integrated multiple sensors and enhanced suspension system to adapt to complex terrain and maintain stable operation.

[0012] 2. Autonomous navigation and positioning technology, using lidar and improved Kalman filter algorithm to achieve accurate autonomous navigation and positioning of the robot in a three-dimensional map.

[0013] 3. Machine vision detection technology, combined with U-Net and YOLOv3 networks, achieves efficient identification of safety hazards in construction scenarios through deep learning optimization.

[0014] 4. Real-time early warning system, integrating sound and light alarms and remote video monitoring, to achieve immediate feedback and processing of detected safety hazards.

[0015] 5. Data storage and analysis, based on the cloud platform, uses data encryption and access control to ensure data security and support efficient analysis and visualization.

[0016] In summary, compared with the prior art, the present invention has the following beneficial effects:

[0017] This machine vision technology combines advanced technologies such as computer vision, image processing, and artificial intelligence to enable real-time monitoring and analysis of complex environments. On construction sites, machine vision can help robots identify and detect potential safety hazards, such as workers not wearing helmets, missing hazardous area signs, and equipment failures.

[0018] Tracked robots, due to their superior terrain adaptability and stability, are ideal for construction site inspections. Equipped with a variety of sensors, such as lidar, color cameras, and depth cameras, the robots can construct three-dimensional environmental maps and achieve autonomous navigation and positioning. Furthermore, the integrated IMU (inertial measurement unit) provides precise posture and motion information, further enhancing the robot's navigation accuracy.

[0019] In machine vision inspection, U-Net neural networks and YOLOv3 networks are two commonly used technologies. U-Net networks are primarily used for image segmentation, accurately extracting the contours of target areas; while YOLOv3 networks excel at object detection, quickly identifying and locating multiple objects. The combination of these two technologies enables robots to efficiently detect various safety hazards at construction sites.

[0020] To provide real-time early warning, the system integrates audio and visual alarms and a remote video monitoring platform. When a safety hazard is detected, the robot immediately alerts on-site personnel through audio and visual alarms and transmits the warning information to the remote monitoring platform for timely response by management personnel.

[0021] Furthermore, the cloud-based data storage and processing platform can store, analyze, and visualize the vast amounts of collected data. Through the user interface, managers can view the safety status of the construction site in real time, receive early warnings, and analyze historical data to develop more effective safety management strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 is a flow chart of the method of the present invention;

[0023] Figure 2 A schematic diagram of the structure of a crawler robot in an embodiment of the present invention, including sensor layout and suspension system design;

[0024] Figure 3 This is a flowchart of the autonomous navigation and positioning technology implementation in an embodiment of the present invention;

[0025] Figure 4 This is a schematic diagram of the application of machine vision detection technology in an embodiment of the present invention, showing the network structure and data processing flow;

[0026] Figure 5 A block diagram of an early warning mechanism system in an embodiment of the present invention, including an integrated approach of sound and light alarms and remote video monitoring;

[0027] Figure 6 This is an architecture diagram of the data storage and processing platform in an embodiment of the present invention, highlighting the data encryption and access control mechanisms;

[0028] Figure 7 This is a schematic diagram of the user interface design in an embodiment of the present invention, showing the customized module and multi-language support interface. DETAILED DESCRIPTION

[0029] Reference Figures 1 to 7 The specific implementation of the machine vision-based robot safety inspection and early warning method for construction sites of construction projects of the present invention is further described.

[0030] A machine vision-based robot safety inspection and early warning method for construction engineering construction sites uses a crawler robot integrated with multiple sensors, utilizes improved autonomous navigation and positioning technology, and combines it with deep learning-driven machine vision detection technology to achieve real-time monitoring of construction sites and identify safety hazards. The specific steps include the following:

[0031] 1. Tracked Robot Body Design: The robot body design takes into account the complex environment of construction sites, including uneven terrain and potential obstacles. The robot body integrates multiple sensors, such as a high-resolution color camera, a precise depth camera, and an IMU, to capture environmental information and provide precise motion data. The suspension system features an enhanced design using shock-resistant materials to adapt to complex terrain and withstand vibration and shock. The robot is also equipped with a high-energy-density lithium-ion battery pack that supports fast charging and hot-swappable battery replacement to ensure long-term operation.

[0032] 2. Autonomous Navigation and Positioning: Autonomous navigation and positioning technology uses LiDAR scanning to construct a three-dimensional map and uses computer vision to analyze significant geometric regions in the image to determine the direction of motion. An improved Kalman filter algorithm estimates the covariance matrix of system noise and measurement noise in real time, adapting to different environments and robot motion states to improve positioning accuracy. SIFT feature point matching technology combined with a locality-sensitive hashing algorithm reduces computational complexity and enables real-time feature matching and pose estimation.

[0033] 3. Machine Vision Inspection: Machine vision inspection technology utilizes U-Net and YOLOv3 neural networks, optimized for construction scenarios. The U-Net network specializes in image segmentation, accurately extracting the outlines of target areas; the YOLOv3 network excels at object detection, rapidly identifying and locating multiple objects. These networks are trained on a specific construction scenario dataset for 2,000 to 5,000 iterations to achieve a detection accuracy of 97% to 99%.

[0034] 4. Real-time Early Warning System: The real-time early warning system integrates audio and visual alarms with a remote video monitoring platform. When machine vision detects a safety hazard, the system immediately alerts on-site personnel through audio and visual alarms and transmits the warning information to the remote monitoring platform for timely action by management. The early warning mechanism automatically adjusts the alarm urgency based on the level of safety risk.

[0035] 5. Data Storage and Analysis: Data storage and analysis is based on a cloud computing platform with data encryption and access control capabilities to ensure data security and integrity. The platform supports efficient data analysis and visualization, enabling managers to view construction site safety status in real time, receive early warning information, and analyze historical data to develop effective safety management strategies.

[0036] 6. User Interface Design: The user interface design focuses on user experience and accessibility, offering an intuitive touch interface and voice command input. The interface style and functional modules are customized to meet the needs of different user groups, including real-time monitoring views, instant warning notifications, and multi-dimensional analysis tools for historical data. The system supports standardized interfaces to connect with existing construction site safety management systems, automatically generates safety reports, and triggers emergency response plans based on detected safety hazards.

[0037] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiment. All technical solutions based on the concept of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A robot safety inspection and early warning method for construction site of a construction project based on machine vision, characterized in that: Using crawler robots, integrating multiple sensors, utilizing improved autonomous navigation and positioning technology, combined with deep learning-driven machine vision detection technology, we can achieve real-time monitoring of construction sites and identification of safety hazards. The specific steps include: a. The crawler robot body is designed to integrate at least the following sensors: high-resolution color camera, precise depth camera, IMU, to adapt to different construction environments and provide accurate motion information; b. Autonomous navigation and positioning, using lidar data to build a three-dimensional map, combining the improved Kalman filter algorithm with adaptive parameter adjustment and SIFT feature point matching technology to achieve precise positioning; c. Machine vision inspection, using U-Net neural networks and YOLOv3 networks optimized for construction scenarios and trained for a specific number of iterations to identify and monitor safety hazards; d. Real-time early warning, integrating an audible and visual alarm system that can automatically adjust the degree of urgency and a remote video monitoring platform to achieve instant feedback and processing of detected safety hazards; e. Data storage and analysis, based on a cloud computing platform with data encryption and access control functions, to ensure data security and integrity, and support efficient analysis and visualization; f. User interface design provides multi-language support, adjusts the level of detail of information display according to user preferences and authority levels, and provides managers with customized real-time monitoring, early warning information reception and historical data analysis functions.

2. The machine vision-based construction site robot safety inspection and early warning method of claim 1, characterized in that: The crawler robot body design includes an enhanced suspension system and shock-proof materials, which are particularly suitable for complex terrain and can withstand the vibrations and shocks common on construction sites.

3. The method for robot safety inspection and early warning of construction site based on machine vision according to claim 1, characterized in that: The improved Kalman filter algorithm in the autonomous navigation and positioning process optimizes the dynamic response by adaptively adjusting the covariance matrix of process noise and measurement noise, as well as the state transfer matrix, thereby achieving higher positioning accuracy and robustness in a variable construction environment.

4. The method for robot safety inspection and early warning of construction site based on machine vision according to claim 1, characterized in that: The U-Net neural network and YOLOv3 network used in the machine vision inspection process are updated through an online learning mechanism to adapt to new construction environments and safety standards. The network is trained for 2,000 to 5,000 iterations on a specific construction scene dataset to achieve a detection accuracy of 97% to 99%.

5. The method for robot safety inspection and early warning of construction site based on machine vision according to claim 1, characterized in that: The early warning mechanism can automatically adjust the urgency of the alarm according to different levels of security risks.

6. The method for robot safety inspection and early warning of construction site based on machine vision according to claim 1, characterized in that: The data storage and analysis platform has data encryption and access control functions to ensure information security.

7. The method for robot safety inspection and early warning of construction site based on machine vision according to claim 1, characterized in that: The user interface provides multi-language support and dynamically adjusts the level of detail of information display and operation permissions based on the user's role, such as project manager, safety supervisor, field worker, and permission level, supporting personalized dashboard configuration and alarm threshold setting.

Images