Pipe gallery construction violation AI identification method and device

Abstract

The invention relates to the field of building construction, and particularly discloses a pipe gallery construction violation AI identification method and device, and the method comprises the steps: information acquisition, AI reasoning, notification and warning, behavior recording, post classification, violation classification, and violation consequence evaluation. The information acquisition unit is used for capturing pictures of a construction site and storing data in a storage medium, and the AI reasoning unit is used for acquiring the data in a wireless transmission mode, analyzing the data through an AI model and acquiring information in the pictures. According to the invention, the personnel are identified and marked, so that the violation behaviors of the personnel and the personnel are matched and recorded, and the marked personnel are emphatically observed in the subsequent recording. According to the method, the violation consequence cost of different working posts in the project is divided, and the marked employees are distributed to posts with low violation consequence cost in the responsibility and capability range, so that the loss of the project is reduced after violation accidents occur.

Description

Technical Field

[0001] The present invention relates to the field of building construction, and more particularly to an AI recognition method and device for illegal construction in utility tunnel construction. Background Art

[0002] A utility tunnel is an underground urban utility tunnel, that is, a tunnel space is built underground in a city, integrating various engineering pipelines such as electricity, communication, heating, water supply and drainage, and radio and television, and is equipped with special inspection openings, hoisting openings and detection systems, and implements unified planning, unified design, unified construction and management, which is an important infrastructure to ensure the operation of the city.

[0003] In the construction of utility tunnels, there may be situations of illegal construction. The illegal construction recognition system can supervise the operations of construction workers in real time to ensure that the construction process meets the specification requirements, thereby improving the construction quality. Therefore, those skilled in the art have provided an AI recognition method and device for illegal construction in utility tunnel construction to solve the problems raised in the above background art. Summary of the Invention

[0004] To solve the above technical problems, the present invention provides an AI recognition method and device for illegal construction in utility tunnel construction, including information acquisition, AI inference, alarm warning, behavior recording, job classification, illegal construction classification, and illegal construction consequence assessment;

[0005] The information acquisition is used to capture the pictures of the construction site and store the data in a storage medium;

[0006] The AI inference obtains data through wireless transmission, analyzes the data through an AI model, and obtains the information in the picture;

[0007] The alarm warning is used to give a notification reminder through an alarm device when an illegal event appears in the picture;

[0008] The behavior recording records the illegal events and stores them in the cloud;

[0009] The job classification centrally classifies the jobs required to complete the project;

[0010] The illegal construction classification classifies the possible illegal events in the project;

[0011] The illegal construction consequence assessment classifies the losses of the project and the injuries of personnel after illegal construction.

[0012] Preferably, the information acquisition includes personnel collection and equipment collection.

[0013] Preferably, the personnel collection includes personnel identity, personnel equipment, personnel behavior, and personnel jobs, and monitors the personnel on the construction site through a camera;

[0014] The device collects data including working hours and equipment operations, and monitors construction equipment at the construction site through cameras.

[0015] Preferably: For the AI inference, a pre-trained AI model is loaded and used for object detection, face recognition, and behavior analysis.

[0016] Preferably: For the alarm warning, when the AI inference detects a violation behavior, it is reminded through an audible and visual alarm device, and the specific violation behavior is played through an audio device.

[0017] Preferably: For the behavior record, the violation behaviors detected by the AI inference are recorded, the personnel are numbered, and the violation records are matched with the personnel numbers.

[0018] Preferably: For the job classification, the jobs with the same functions in the project are classified, and the number of the same jobs is recorded.

[0019] Preferably: For the violation classification, it is classified according to the risk level caused by the violation behavior in the same job.

[0020] Preferably: In the evaluation of the consequences of violations, the project losses include economic losses and time losses.

[0021] The present invention also provides an AI recognition device for construction violations in pipe galleries, including:

[0022] A monitoring device, multiple groups of cameras installed in different areas of the site, a storage medium for storing data, and a communication module for wireless transmission;

[0023] A video processing device, which maintains wireless transmission with the monitoring device and analyzes and processes the transmitted data;

[0024] An audible and visual alarm device, including a buzzer, an alarm lamp, and a speaker.

[0025] The technical effects and advantages of the present invention:

[0026] 1. The present invention identifies and marks personnel, thereby matching and recording the violation behaviors of personnel with the personnel, and in subsequent records, key observations are made on the marked personnel.

[0027] 2. The present invention divides the sizes of the consequences of violations of different job positions in the project, and assigns the marked employees to the positions with small consequences of violations within their responsibilities and capabilities, so as to reduce the losses of the project after a violation accident occurs. Description of the Drawings

[0028] Figure 1 It is a schematic structural diagram provided by this application;

[0029] Figure 2 is a schematic structural diagram of the violation behavior provided by the present application;

[0030] Figure 3 is a schematic structural diagram of the confidence acquisition provided by the present application; Detailed implementation manners

[0031] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", etc. indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention.

[0032] In the description of the present invention, the meaning of "a plurality of" is two or more, unless otherwise specifically defined.

[0033] In the description of the present invention, it should be noted that unless otherwise clearly specified and limited, the terms "installed", "provided with", "sheathed / connected", "connected", etc. should be understood in a broad sense. For example, "connected" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0034] Please refer to Figures 1 to 3 , in this embodiment, a method and device for AI recognition of violations in utility tunnel construction are provided, including information acquisition, AI inference, alarm warning, behavior recording, job classification, violation classification, and violation consequence assessment;

[0035] The information acquisition is used to capture the picture of the construction site and store the data in a storage medium;

[0036] The AI inference obtains data through a wireless transmission method, analyzes the data through an AI model, and obtains the information in the picture;

[0037] S1. Video stream acquisition and decoding

[0038] Video stream acquisition: The AI system can acquire video streams in various ways, including directly connecting to a camera, pulling video streams through network protocols (such as RTSP, RTMP), or processing video files stored locally.

[0039] Video decoding: Decoding is the process of converting a video stream from a compressed format to a sequence of frames that can be processed. This process can be accomplished using hardware acceleration (such as GPUs or dedicated decoding chips) or software decoding tools (such as FFmpeg).

[0040] S2. Frame extraction and preprocessing

[0041] Frame rate control: Extract frames at a specific frame rate according to requirements, such as 1 frame per second or 30 frames per second, to balance real-time performance and computational resources.

[0042] Resolution adjustment: Adjust the resolution of the frames to meet the input requirements of the AI model, which helps improve the inference speed.

[0043] S3. AI processing

[0044] Model loading: Load pre-trained AI models, such as YOLO, SSD, ResNet, etc., which can be used for tasks such as object detection, face recognition, and behavior analysis.

[0045] Inference acceleration: Use hardware acceleration technologies (such as NVIDIA TensorRT, Intel OpenVINO) to boost the inference speed and ensure real-time performance.

[0046] Real-time analysis: Analyze each frame to identify objects, behaviors, or events in the video.

[0047] S4. Result processing and visualization

[0048] Annotation and visualization: Annotate the detection results on the video frames, such as bounding boxes, labels, etc., for intuitive display of the analysis results.

[0049] Event triggering: Trigger alarms or other operations based on the analysis results, such as sending notifications or storing video clips.

[0050] S5. Optimization and tuning

[0051] Model optimization: Optimize the model through techniques such as quantization and pruning to improve the inference speed.

[0052] Resource management: Reasonably allocate CPU, GPU, and memory resources to ensure the stable operation of the system.

[0053] Regarding the above-mentioned warning, when a violation event appears in the picture, it is notified and reminded through an alarm device;

[0054] The behavior record records the violation events and stores them in the cloud;

[0055] The job classification centrally classifies the jobs required to complete the project;

[0056] The violation classification classifies the possible violation events in the project;

[0057] The evaluation of the consequences of violations grades the losses of the project and the injuries of personnel after violations.

[0058] In a further embodiment, the information acquisition includes personnel collection and equipment collection;

[0059] The personnel collection includes personnel identity, personnel equipment, personnel behavior, and personnel positions, and monitors the personnel at the construction site through cameras;

[0060] The equipment collection includes working hours and equipment operation, and monitors the construction equipment at the construction site through cameras.

[0061] In a further embodiment, the AI inference loads a pre-trained AI model and is used for object detection, face recognition, and behavior analysis.

[0062] In a further embodiment, the alarm warning, when the AI inference detects a violation behavior, reminds through an audible and visual alarm device and plays the specific violation behavior through an audio device.

[0063] In a further embodiment, the behavior record records the violation behaviors detected by the AI inference, numbers the personnel, and matches the violation records with the personnel numbers.

[0064] In a further embodiment, the job classification classifies the jobs with the same functions in the project and records the number of the same jobs.

[0065] In a further embodiment, the violation classification classifies according to the danger level caused by the occurrence of violation behaviors among the same jobs.

[0066] In a further embodiment, the project losses in the evaluation of the consequences of violations include economic losses and time losses.

[0067] In this embodiment, an AI recognition device for violations in pipe gallery construction is provided, including:

[0068] A monitoring device, multiple groups of cameras installed in different areas of the site, a storage medium for storing data, and a communication module for wireless transmission;

[0069] The screen processing device maintains wireless transmission with the monitoring device and analyzes and processes the transmitted data;

[0070] The sound and light alarm device includes a buzzer, an alarm light, and a speaker.

[0071] Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art and related fields without creative efforts shall fall within the scope of protection of the present invention. The structures, devices, and operation methods not specifically described and explained in the present invention shall be implemented by conventional means in the art without special description and limitation.

Claims

1. An AI recognition method and device for construction violations in utility tunnels, characterized in that, It includes information acquisition, AI reasoning, alarm warning, behavior recording, job classification, violation classification, and violation consequence assessment; The information acquisition is used to capture the images of the construction site and store the data in a storage medium; The AI reasoning obtains data through wireless transmission, analyzes the data through an AI model, and obtains the information in the images; The alarm warning is used to give a notice and reminder through an alarm device when a violation event appears in the images; The behavior recording records the violation events and stores them in the cloud; The job classification centrally classifies the jobs required to complete the project; The violation classification classifies the possible violation events in the project; The violation consequence assessment classifies the project losses and personnel injuries after a violation into different levels.

2. The AI recognition method for construction violations in utility tunnels according to claim 1, wherein, The information acquisition includes personnel collection and equipment collection; The personnel collection includes personnel identity, personnel equipment, personnel behavior, and personnel jobs, and monitors the personnel at the construction site through cameras; The equipment collection includes working hours and equipment operation, and monitors the construction equipment at the construction site through cameras.

3. The AI recognition method for construction violations in utility tunnels according to claim 1, wherein The AI reasoning loads a pre-trained AI model and is used for object detection, face recognition, and behavior analysis.

4. The AI recognition method for construction violations in utility tunnels according to claim 1, characterized in that, The alarm warning, when the AI reasoning detects a violation behavior, gives a reminder through an audible and visual alarm device and plays the specific violation behavior through an audio device.

5. The AI recognition method for construction violations in utility tunnels according to claim 1, wherein, The behavior recording records the violation behaviors detected by the AI reasoning, numbers the personnel, and matches the violation records with the personnel numbers.

6. The AI recognition method for construction violations in utility tunnels according to claim 1, characterized in that, The job classification classifies the jobs with the same functions in the project and records the number of the same jobs.

7. The AI recognition method for construction violations in utility tunnels according to claim 1, characterized in that, The violation classification classifies according to the risk level caused by the violation behavior among the same jobs.

8. The AI recognition method for construction violations in utility tunnels according to claim 1, characterized in that, In the violation consequence assessment, the project losses include economic losses and time losses.

9. An AI recognition device for detecting violations in utility tunnel construction, characterized in that, It includes: A monitoring device, multiple groups of cameras installed in different areas of the site, a storage medium for storing data, and a communication module for wireless transmission; An image processing device, which maintains wireless transmission with the monitoring device and analyzes and processes the transmitted data; An audible and visual alarm device, including a buzzer, an alarm lamp, and an audio.

Images