Intelligent construction site management method for real-time potential safety hazard checking based on AI algorithm

Abstract

The invention discloses an intelligent construction site management method for real-time potential safety hazard investigation based on an AI algorithm, and relates to the technical field of construction site safety management, and the method comprises the steps: image collection: setting a plurality of high-resolution cameras in a construction region, and shooting and collecting images in the construction region; preprocessing the image, removing useless pictures, and unifying the size and format of the image; training an AI model, and constructing a potential safety hazard model; image analysis and identification: inputting the preprocessed image into a potential safety hazard model for real-time analysis and identification; a result is fed back, and when potential safety hazards are recognized, the system automatically triggers an alarm; and data storage and management: storing all the collected images and analysis results, and regularly generating a report. According to the invention, accidents can be effectively prevented, the personal safety of workers is guaranteed, the workload of manual inspection is reduced, managers can quickly respond to potential risk points, the whole monitoring process is simplified, and construction site safety management is more efficient and convenient.

Description

Technical Field

[0001] The present invention belongs to the technical field of construction site safety management, and particularly relates to a smart construction site management method for real-time safety hazard detection based on AI algorithms. Background Art

[0002] In the existing safety management of construction sites, on-site problems are generally discovered by means of manual on-site inspections or manual supervision in the monitoring room. The existing methods are not only time-consuming and laborious, but also cannot achieve all-round and round-the-clock safety inspections. Moreover, the recognition accuracy of the existing technologies is unreliable, real-time monitoring cannot be achieved, and at the same time, it is impossible to transmit pictures of safety problems and safety hazard problems across platforms. Therefore, we propose a smart construction site management method for real-time safety hazard detection based on AI algorithms. Summary of the Invention

[0003] The purpose of the present invention is to solve the above-mentioned drawbacks in the existing technology, and to propose a smart construction site management method for real-time safety hazard detection based on AI algorithms. This smart construction site management method for real-time safety hazard detection based on AI algorithms can effectively prevent accidents, ensure the personal safety of workers, reduce the workload of manual inspections, enable managers to quickly respond to potential risk points, simplify the entire monitoring process, and make construction site safety management more efficient and convenient.

[0004] In order to achieve the above purpose, the present invention adopts the following technical solutions:

[0005] Design a smart construction site management method for real-time safety hazard detection based on AI algorithms, including the following steps:

[0006] Step 1, Image acquisition: Set up multiple high-resolution cameras in the construction area and capture images in the construction area.

[0007] Step 2, Image preprocessing: Remove useless pictures, and unify the image size and format.

[0008] Step 3, AI model training: Build a safety hazard model.

[0009] Step 4, Image analysis and recognition: Input the preprocessed images in Step 2 into the safety hazard model for real-time analysis and recognition.

[0010] Step 5, Result feedback: When a safety hazard is identified, the system automatically triggers an alarm, and pushes the image with the safety hazard and the tracking information to the smart construction site management platform.

[0011] Step 6, Data storage and management: Save all the captured images and analysis results, and generate reports regularly.

[0012] Further, in step 1, the monitoring ranges of the multiple high-resolution cameras set cover the entire construction area, forming a continuous modular monitoring area, and images within the construction area are captured by means of taking regular shots or triggering shots based on motion detection.

[0013] Further, in step 2, the specific method for removing useless pictures includes the following steps:

[0014] Step 21, set a reference image for each fixed camera area;

[0015] Step 22, compare and analyze the images collected in the corresponding area with the reference image, remove the same images, and retain the images with changes;

[0016] Step 23, conduct a secondary comparison on the images with changes, remove the same areas in the images, and retain the different areas to form a solid-color background image containing the changed areas;

[0017] Step 24, screen out the solid-color background images of the changed areas containing people through an image and person recognition algorithm.

[0018] Further, in step 3, the steps for constructing a safety hazard model are as follows:

[0019] Step 31, dataset construction: collect images containing various types of safety hazards as the training dataset;

[0020] Step 32, annotation work: annotate the images with safety hazards in the dataset;

[0021] Step 33, model selection and training: use deep learning technology to train the model to identify safety hazards.

[0022] Further, in step 4, evaluate the risk level of the analysis and identification results.

[0023] Further, in step 5, the tracking information at least includes the activity paths of the people with safety hazards within the construction area.

[0024] Further, the determination of the activity paths includes the following steps:

[0025] Step 51, conduct feature recognition on the images containing safety hazards;

[0026] Step 52, calibrate the modular monitoring areas of the images with the same feature safety hazards according to the time series as t1, t2, t3... tn;

[0027] Step 53, connect the calibrated modular monitoring areas of t1, t2, t3... tn according to the time axis to form the activity paths of the people with safety hazards within the construction area.

[0028] A smart construction site management method based on real-time safety hazard detection using AI algorithms has the following beneficial effects:

[0029] (1) This invention can improve the safety of the construction area: By promptly detecting and reporting safety hazards at the construction site, accidents can be effectively prevented, ensuring the personal safety of workers.

[0030] (2) This invention can enhance the management efficiency of the construction area. Automated identification and alarm reduce the workload of manual inspections, enabling managers to quickly respond to potential risk points.

[0031] (3) This invention can reduce operating costs, reducing the time and labor costs required for manual patrols. At the same time, insurance costs and compensation costs are indirectly reduced due to the decreased accident rate.

[0032] (4) This invention can simplify the monitoring process. The system automatically completes tasks such as image acquisition, analysis, and report generation, simplifying the entire monitoring process and making construction site safety management more efficient and convenient.

[0033] (5) This invention has a low failure rate. Due to the adoption of mature AI technology and high-quality hardware facilities, the system has good stability and a low failure rate.

[0034] (6) This invention has strong reliability. It can achieve 24-hour uninterrupted monitoring, is not affected by weather conditions, ensures the effectiveness of all-weather monitoring, and can map the movement paths of safety hazard personnel within the construction area, facilitating the search for such personnel. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. They are used in conjunction with the embodiments of the present invention to explain the present invention, but do not limit the present invention. In the drawings:

[0036] Figure 1 is a flowchart of the present invention;

[0037] Figure 2 is a schematic diagram showing the formation of movement paths of safety hazard personnel within the construction area in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0038] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0039] Now, in combination with the accompanying drawings of the specification, the structural features of the present invention will be described in detail.

[0040] See Figure 1 - Figure 2 , a smart construction site management method for real-time safety hazard detection based on AI algorithms, comprising the following steps:

[0041] S1. Image acquisition: Set multiple high-resolution cameras in the construction area and capture images of the construction area.

[0042] In order to more comprehensively collect images in the construction area, among them, the monitoring ranges of the multiple high-resolution cameras set cover the entire construction area, forming a continuous modular monitoring area, and images in the construction area are captured by means of regular shooting or shooting triggered by motion detection.

[0043] S2. Image preprocessing: Remove useless pictures, and unify the image size and format.

[0044] Among them, the specific method for removing useless pictures includes the following steps:

[0045] Step 21: Set a reference image for each fixed camera area;

[0046] Step 22: Compare and analyze the images collected in the corresponding area with the reference image, remove the same images, and retain the images with changes;

[0047] Step 23: Perform a secondary comparison on the images with changes, remove the same areas in the images, and retain the different areas to form a pure-color background image containing the changed areas;

[0048] Step 24: Screen out the pure-color background images of the changed areas containing people through the image and person recognition algorithm.

[0049] Image preprocessing can reduce the difficulty of subsequent image processing, so as to effectively clean, standardize, enhance, etc. the preprocessing of the collected image data for subsequent processing, so as to improve the recognition accuracy. Ensure the ability to complete the analysis of a large number of images in a short time and make decisions quickly in the future.

[0050] S3. AI model training: Build a safety hazard model.

[0051] Among them, the steps for building a safety hazard model are as follows:

[0052] Step 31. Dataset construction: Collect images containing various types of safety hazards as the training dataset;

[0053] Step 32. Annotation work: Annotate the images with safety hazards in the dataset;

[0054] Step 33, Model Selection and Training: Use deep learning techniques, such as Convolutional Neural Network (CNN), to train a model to identify potential safety hazards.

[0055] S4, Image Analysis and Recognition: Input the preprocessed images in S2 into the safety hazard model for real-time analysis and recognition. Evaluate the risk level of the analysis and recognition results.

[0056] S5, Result Feedback: When a safety hazard is identified, the system automatically triggers an alarm and pushes the image with the safety hazard and the tracking information to the intelligent construction site management platform. Specifically, push the photo and detailed information of the safety hazard to the intelligent construction site management platform through the API interface. Design a reasonable information push strategy to ensure that the alarm information can be transmitted to the relevant responsible persons in a timely and accurate manner.

[0057] The tracking information includes at least the activity path of the person with the safety hazard within the construction area. The determination of the activity path includes the following steps:

[0058] Step 51, Perform feature recognition on the image containing the safety hazard;

[0059] Step 52, Calibrate the modular monitoring areas of the images with the same feature safety hazards in time series as t1, t2, t3... tn;

[0060] Step 53, Connect the calibrated modular monitoring areas of t1, t2, t3... tn in series according to the time axis to form the activity path of the person with the safety hazard within the construction area.

[0061] S6, Data Storage and Management: Save all the collected images and analysis results, and generate reports regularly. Save all the collected images and analysis results for later auditing and tracing, and generate safety hazard reports regularly to help the management understand the safety status of the construction site.

[0062] The intelligent construction site management method for real-time safety hazard investigation based on AI algorithms, which is a method of seamlessly integrating AI recognition technology into the existing construction site management system. On the one hand, by promptly discovering and reporting safety hazards at the construction site, it can effectively prevent accidents from occurring and ensure the personal safety of workers. It can enhance the management efficiency of the construction area. The automated identification and alarm reduce the workload of manual inspections, enabling managers to quickly respond to potential risk points. On the other hand, during use, it can reduce operating costs, reducing the time and labor costs required for manual patrols. At the same time, the insurance costs and compensation costs are indirectly reduced due to the decrease in the accident rate. It can simplify the monitoring process. The system automatically completes tasks such as image acquisition, analysis, and report generation, simplifying the entire monitoring process and making the construction site safety management more efficient and convenient. In addition, due to the adoption of mature AI technology and high-quality hardware facilities, the system has good stability, low failure rate, and strong reliability. It can achieve 24-hour uninterrupted monitoring, is not affected by weather conditions, ensures the effectiveness of all-weather monitoring, and can formulate the activity paths of safety hazard personnel within the construction area, facilitating the search for safety hazard personnel.

[0063] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions recorded in the foregoing embodiments or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A smart construction site management method based on AI algorithm for real-time safety hazard detection, characterized in that: The steps include: Step 1: Image acquisition: multiple high-resolution cameras are set up in the construction area to capture images of the construction area; Step 2: Image preprocessing: removing useless images and unifying image size and format; Step 3: AI model training to build a safety hazard model; Step 4: Image analysis and recognition: input the image preprocessed in step 2 into the safety hazard model for real-time analysis and recognition; Step 5: Result feedback: When a potential safety hazard is identified, the system automatically triggers an alarm and pushes the image and tracking information of the potential safety hazard to the smart construction site management platform; Step 6: Data storage and management: save all collected images and analysis results, and generate reports regularly.

2. According to claim 1, a smart construction site management method based on AI algorithm for real-time safety hazard detection is characterized in that: In step 1, the monitoring range of multiple high-resolution cameras covers the entire construction area to form a continuous modular monitoring area, and images in the construction area are captured by timed shooting or shooting triggered by motion detection.

3. According to claim 1, a smart construction site management method based on AI algorithm for real-time safety hazard detection is characterized in that: In step 2, the specific method of removing useless images includes the following steps: Step 21, setting a reference image for each fixed camera area; Step 22, comparing and analyzing the images collected in the corresponding area with the reference image, removing the identical images and retaining the images with changes; Step 23, performing a secondary comparison on the changed image, removing the same areas in the image, retaining the different areas, and forming a pure color background image containing the changed areas; Step 24, using an image character recognition algorithm to filter out the pure color background image containing the changing area of ​​the character.

4. According to claim 1, a smart construction site management method based on AI algorithm for real-time safety hazard detection is characterized in that: In step 3, the steps to build the safety hazard model are as follows: Step 31, data set construction: collect images containing various types of safety hazards as training data sets; Step 32: Labeling: labeling the images with potential safety hazards in the data set; Step 33: Model selection and training: Use deep learning technology to train the model to identify safety hazards.

5. According to claim 1, a smart construction site management method based on AI algorithm for real-time safety hazard detection is characterized in that: In step 4, the risk level of the analysis and identification results is evaluated.

6. According to claim 1, a smart construction site management method based on AI algorithm for real-time safety hazard detection is characterized in that: In step 5, the tracking information includes at least the activity path of the person who poses a safety hazard in the construction area.

7. The intelligent construction site management method based on AI algorithm for real-time safety hazard detection according to claim 7 is characterized in that: The determination of the activity path includes the following steps: Step 51, performing feature recognition on images containing potential safety hazards; Step 52, calibrating the modular monitoring areas containing the safety hazard images with the same characteristics according to the time series t1, t2, t3...tn; Step 53, connecting the modular monitoring areas calibrated at t1, t2, t3, ..., tn in series according to the time axis to form an activity path of people with safety hazards in the construction area.

Images