Detection method, system and equipment for low hanging and high use of safety belt in high-place operation and medium

Abstract

The invention discloses a method, a system, equipment and a medium for detecting low hanging and high use of a high-place operation safety belt, and relates to the field of operation safety detection. The method comprises the steps of obtaining a to-be-detected image; preprocessing the to-be-detected image to obtain a to-be-detected enhanced channel image; detecting the to-be-detected enhanced channel image to obtain a first marking result and a second marking result; segmenting the to-be-detected enhanced channel image with the first marking result and the second marking result to obtain an operator marking area and a safety belt marking area; the part, coinciding with the operator marking area, in the safety belt marking area is removed, and a target marking area is obtained; determining a first height value and a second height value according to the operator marking area, and determining a third height value according to the target marking area; and according to the first height value, the second height value and the third height value, determining whether the safety belt is hung at a low position and used at a high position in the target operation site. According to the invention, the detection efficiency and accuracy of safety belt use specifications can be improved.

Description

Technical Field

[0001] This invention relates to the field of work safety testing, and in particular to a testing method, system, equipment, and medium for detecting low-hanging, high-use safety belts for high-altitude operations. Background Technology

[0002] In power maintenance sites, there are numerous instances of work at height and overlapping operations, and the work environments are complex, with intricate relative positions of personnel, equipment, and tools. According to power company safety regulations, to prevent falls during height work, workers must wear safety harnesses. However, during operations, workers frequently need to change positions to complete different procedures, requiring multiple changes to the safety harness attachment points. The process of putting on a safety harness is lengthy and involves numerous preparation steps, and its proper use is often overlooked by personnel performing short-duration, low-load tasks. Extensive accident analysis has revealed a typical violation involving safety harnesses: using them improperly with the harness positioned too low.

[0003] Any work performed at a height of 2 meters or more above the fall reference plane is considered work at height. Fall protection is mandatory for work at height. In work environments lacking fall protection, safety harnesses must be worn when working at a height exceeding 1.5 meters. Generally, before wearing a safety harness, it should first be inspected: check for deterioration of the harness, cracks in the shackles, and proper spring action. Secondly, in work at height where there is no fixed anchor point for the safety harness, a steel wire rope of appropriate strength or other methods should be used. The correct method for using a safety harness is "high anchor, low use," meaning the harness is attached high, and the worker works below the anchor point. This high anchor, low use method is a safe and scientifically sound method, reducing the actual impact distance in the event of a fall. Conversely, the low anchor, high use method—attaching the harness low while working above—is a very unsafe method. When a worker falls, the fall will pass through the anchor point, increasing the actual impact distance. Both the worker and the rope will be subjected to a greater impact load, which may result in a fatal accident in severe cases.

[0004] Fall injuries are divided into two types: impact injuries and strike injuries during the fall. Impact injuries occur when the fall comes to a stop and the momentum is converted into impulse, causing injury to the faller's body. Strike injuries during the fall occur when the faller comes into contact with other structures. Improperly using safety belts with a low attachment point increases the impact load on the faller by 42%. This practice significantly increases the accident risk for workers. Intelligent monitoring of the entire construction process can regulate worker actions.

[0005] In power work sites, traditional safety belt compliance checks are often conducted manually, with safety inspectors providing on-site supervision. However, due to factors such as obstructed vision, this manual inspection method is not only inefficient but also unable to address blind spots. Summary of the Invention

[0006] The purpose of this invention is to provide a method, system, equipment, and medium for detecting the low-hanging and high-use of safety belts for high-altitude operations, so as to improve the efficiency and accuracy of detecting safety belt usage standards.

[0007] To achieve the above objectives, the present invention provides the following solution:

[0008] A method for detecting the low-hanging, high-use of safety belts for working at heights, the method comprising:

[0009] Acquire the image to be detected; the image to be detected is acquired by a camera located above the working plane of the target work site, facing the workers and safety belts;

[0010] An image enhancement algorithm is used to preprocess the image to be detected to obtain the enhanced channel image to be tested;

[0011] An object detection model is used to detect the image of the enhancement channel to be tested, and a first labeling result and a second labeling result are obtained. The first labeling result is the worker area label and the safety belt area label in the image of the enhancement channel to be tested. The second labeling result is the worker boundary label and the safety belt boundary label in the image of the enhancement channel to be tested. The object detection model is trained based on the Mask R-CNN network.

[0012] An image segmentation algorithm is used to segment the image of the enhanced channel to be tested, which contains the first marking result and the second marking result, to obtain the worker marking area and the safety belt marking area;

[0013] Remove the portion of the safety belt marking area that overlaps with the worker marking area to obtain the target marking area;

[0014] A first height value and a second height value are determined based on the worker-marked area, and a third height value is determined based on the target marked area; the first height value is the highest pixel height of the worker-marked area; the second height value is the lowest pixel height of the worker-marked area; and the third height value is the highest pixel height of the target marked area.

[0015] Based on the first height value, the second height value, and the third height value, determine whether there is a phenomenon of low-hanging and high-use of safety belts at the target work site.

[0016] Optionally, determining whether the target work site has a situation where the safety belt is used at a lower position than its intended purpose based on the first height value, the second height value, and the third height value specifically includes:

[0017] The pixel height of the worker area is determined based on the first height value and the second height value;

[0018] The pixel height of the seat belt attachment point is determined based on the third height value and the second height value;

[0019] Determine whether the pixel height of the safety belt attachment point is less than a set multiple of the pixel height of the worker area, and obtain the determination result; the set multiple is 0.75 times.

[0020] If the judgment result is yes, then it is determined that there is a phenomenon of low-hanging and high-use of safety belts at the target work site;

[0021] If the judgment result is negative, then it is determined that there is no phenomenon of low-hanging and high-use of safety belts at the target work site.

[0022] Optionally, the step of employing an image enhancement algorithm to preprocess the image to be detected to obtain the image of the enhanced channel to be tested specifically includes:

[0023] The image to be detected is subjected to RGB channel separation, and the intensity channel image is determined based on the separated blue channel image, green channel image, and red channel image;

[0024] A multi-scale retinal enhancement algorithm is used to enhance the intensity channel image to obtain an initial enhanced image;

[0025] According to the set magnification factor, the pixel values ​​of each channel in the initial enhanced image are magnified to obtain the image of the enhanced channel to be tested.

[0026] Optionally, the target detection model includes a first target detection layer and a second target detection layer connected in series;

[0027] The first target detection layer is used to detect the image of the enhanced channel to be tested and obtain a first labeling result;

[0028] The second target detection layer is used to detect the image of the enhanced channel to be tested with the first labeling result to obtain the second labeling result.

[0029] Optionally, the step of employing an image segmentation algorithm to segment the enhanced channel image under test, which contains the first and second labeling results, to obtain the worker labeling region and the safety belt labeling region, specifically includes:

[0030] The Otsu threshold segmentation algorithm is used to segment the image of the enhanced channel to be tested, which contains the first and second labeling results, to obtain the worker area and the safety belt area.

[0031] Optionally, the step of determining the pixel height of the worker area based on the first height value and the second height value is specifically formulated as follows:

[0032] h C =y C -y' C

[0033] Among them, h C The pixel height of the worker area, y C For the first height value, y' C This is the second altitude value.

[0034] Optionally, the step of determining the pixel height of the seatbelt attachment point based on the third height value and the second height value is specifically formulated as follows:

[0035] y' E =y E -y' C

[0036] Among them, y' E For the pixel height of the seatbelt attachment point, y E The third height value, y' C This is the second altitude value.

[0037] A detection system for low-hanging, high-use safety belts used in high-altitude operations, the system comprising:

[0038] The image acquisition module is used to acquire the image to be detected; the image to be detected is acquired by a camera located above the working plane of the target work site, facing the workers and safety belts;

[0039] The image enhancement module is used to preprocess the image to be detected using an image enhancement algorithm to obtain the enhanced channel image to be tested;

[0040] The target detection module is used to detect the image of the enhanced channel to be tested using a target detection model, and obtain a first labeling result and a second labeling result; the first labeling result is the worker area label and the safety belt area label in the image of the enhanced channel to be tested; the second labeling result is the worker boundary label and the safety belt boundary label in the image of the enhanced channel to be tested.

[0041] The image segmentation module is used to segment the image of the enhanced channel to be tested, which contains the first marking result and the second marking result, using an image segmentation algorithm to obtain the worker marking area and the safety belt marking area;

[0042] The region determination module is used to remove the portion of the safety belt marking region that overlaps with the worker marking region to obtain the target marking region;

[0043] A height determination module is used to determine a first height value and a second height value based on the worker-marked area, and to determine a third height value based on the target marked area; the first height value is the highest pixel height of the worker-marked area; the second height value is the lowest pixel height of the worker-marked area; and the third height value is the highest pixel height of the target marked area.

[0044] The safety detection module is used to determine whether there is a phenomenon of low-hanging and high-use of safety belts at the target work site based on the first height value, the second height value, and the third height value.

[0045] An electronic device includes a memory and a processor, the memory storing a computer program, and the processor running the computer program to enable the electronic device to perform the above-described detection method for low-hanging, high-use safety belts in high-altitude operations.

[0046] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described detection method for low-mounted, high-use safety belts for high-altitude operations.

[0047] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0048] This invention acquires an image to be inspected and sequentially performs image enhancement, target detection, and image segmentation processing to determine the target marking area within the worker marking area and the safety belt marking area. Based on the pixel heights of the highest and lowest points in these areas, it determines whether a safety belt is being used at a lower position than its intended function at the work site. Compared to manual inspection, this invention enables real-time automatic detection, improving efficiency and avoiding low accuracy issues caused by factors such as obstructed vision. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 A flowchart of the detection method for low-hanging and high-use safety belts for high-altitude operations provided by the present invention;

[0051] Figure 2 A schematic diagram of a power operation site provided for an embodiment of the present invention;

[0052] Figure 3 This is a schematic diagram of the worker marking area and the safety belt marking area provided in an embodiment of the present invention;

[0053] Figure 4 A schematic diagram of the target marking region provided in an embodiment of the present invention;

[0054] Figure 5 A block diagram of the detection system for low-mounted, high-use safety belts for high-altitude operations provided by the present invention. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] In recent years, with the popularization of intelligent monitoring equipment for on-site operations, smart cameras have gradually replaced safety inspectors to investigate various risk factors at the work site.

[0057] Based on this, the purpose of this invention is to provide a detection method, system, equipment, and medium for detecting the low-hanging and high-use of safety belts for high-altitude operations, so as to improve the detection efficiency and accuracy of safety belt usage standards.

[0058] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0059] Example 1

[0060] like Figure 1 As shown, the present invention provides a method for detecting the low-hanging, high-use of safety belts for high-altitude operations, the method comprising:

[0061] Step 101: Acquire the image to be detected; the image to be detected is acquired by a camera located above the working plane of the target work site, facing the workers and safety belts.

[0062] Step 102: The image to be detected is preprocessed using an image enhancement algorithm to obtain the image of the enhanced channel to be tested. Step 102 specifically includes:

[0063] Step 102.1: Perform RGB channel separation on the image to be detected, and determine the intensity channel image based on the separated blue channel image, green channel image and red channel image.

[0064] Step 102.2: Use a multi-scale retinal enhancement algorithm to enhance the intensity channel image to obtain an initial enhanced image.

[0065] Step 102.3: Magnify the pixel values ​​of each channel in the initial enhanced image according to the set magnification factor to obtain the image of the enhanced channel to be tested. The set magnification factor is determined by the pixel with the largest pixel value in the image to be tested.

[0066] Step 103: Using a target detection model, the image to be enhanced is detected to obtain a first labeling result and a second labeling result; the first labeling result is the worker area label and the safety belt area label in the image to be enhanced; the second labeling result is the worker boundary label and the safety belt boundary label in the image to be enhanced; the target detection model is trained based on the Mask R-CNN network.

[0067] Specifically, the target detection model includes a first target detection layer and a second target detection layer connected in series; the first target detection layer is used to detect the image of the enhanced channel to be tested to obtain a first labeling result; the second target detection layer is used to detect the image of the enhanced channel to be tested with the first labeling result to obtain a second labeling result.

[0068] Step 104: Using an image segmentation algorithm, the image of the enhancement channel to be tested, which contains the first and second labeling results, is segmented to obtain the worker-marked region and the safety belt-marked region. Preferably, the image segmentation algorithm is the Otsu thresholding algorithm.

[0069] Step 105: Remove the portion of the safety belt marking area that overlaps with the worker marking area to obtain the target marking area.

[0070] Step 106: Determine a first height value and a second height value based on the worker-marked area, and determine a third height value based on the target marked area; the first height value is the highest pixel height of the worker-marked area; the second height value is the lowest pixel height of the worker-marked area; and the third height value is the highest pixel height of the target marked area.

[0071] Step 107: Determine whether the target work site exhibits a phenomenon of low-hanging, high-use of safety belts based on the first height value, the second height value, and the third height value. Step 107 specifically includes:

[0072] Step 107.1: Determine the pixel height of the worker area based on the first height value and the second height value, using the following formula:

[0073] h C =y C -y' C

[0074] Among them, h C The pixel height of the worker area, y C For the first height value, y' C This is the second altitude value.

[0075] Step 107.2: Determine the pixel height of the seat belt attachment point based on the third height value and the second height value, using the following formula:

[0076] y' E =y E -y' C

[0077] Among them, y' E For the pixel height of the seatbelt attachment point, y E The third height value, y' C This is the second altitude value.

[0078] Step 107.3: Determine whether the pixel height of the safety belt attachment point is less than a set multiple of the pixel height of the worker area, and obtain the determination result; the set multiple is 0.75 times.

[0079] Step 107.4: If the judgment result is yes, then it is determined that the target work site has a phenomenon of low-mounted and high-used safety belts; if the judgment result is no, then it is determined that the target work site does not have a phenomenon of low-mounted and high-used safety belts.

[0080] Furthermore, the detection method further includes:

[0081] Step 108: When the target work site has a situation where the safety belt is used at a low position but not at a high position, an alarm signal is generated to alert the workers and managers so that the workers can use the safety belt in a standardized manner.

[0082] The following provides a specific embodiment to illustrate the execution process of the present invention in detail.

[0083] 1) such as Figure 2 As shown, first, a camera is installed on the work surface, and the camera height should be parallel to the main work area.

[0084] 2) During the operation, the camera is used to capture the workers' working status at a rate of 1 frame per second.

[0085] 3) The image is preprocessed using a multi-scale Retinex-enhanced image stitching algorithm. The basic process of this algorithm is as follows:

[0086] (a): Separation of the original Figure 3 The image consists of three channels: IB (blue channel), IG (green channel), and IR (red channel), and the intensity channel image I is calculated.

[0087] (b): Apply the Multi-Scale Retinex (MSR) algorithm to the intensity channel image I, and denote the enhanced image as IMSR.

[0088] (c): Let the pixel value of the i-th pixel in the original image be the maximum value of the pixel values ​​of the three channels. Calculate the magnification factor of each channel at this point, and then magnify the pixels of the three channels respectively until the entire image is magnified.

[0089] 4) The Mask R-CNN algorithm is used to detect defects in the enhanced channel image. The task is to identify the workers and safety belts in the image, forming the first labeling result A. This step specifically includes:

[0090] (a): Input the image obtained in step 3) into the pre-trained ResNeXt layer to obtain the corresponding feature map, i.e., to label the workers in the image. The specific training process of the ResNeXt layer is as follows:

[0091] (i): The training sample consists of more than 1,000 images of workers wearing safety belts, in which the safety belts, workers, and typical power system equipment need to be pre-labeled.

[0092] (ii): This training step only needs to be performed once. In field applications, the trained ResNeXt layer has the ability to mark safety belts, workers and typical power system equipment in images.

[0093] (b): Set a predetermined number of Region of Interest (ROI) for each point in this feature map to obtain multiple candidate ROIs.

[0094] (c): These candidate ROIs are fed into the RPN network layer for binary classification to determine which areas are foreground and which are background. That is, in the image from step 3, workers and safety belts are marked as foreground, and electrical equipment as background. The ResNeXt layer and the RPN network layer together constitute the first target detection layer.

[0095] 5) Align the ROI obtained in step 4 with the image in step 3 to form a labeled region. Use the Mask R-CNN algorithm to detect the enhanced channel image and further mark the boundaries of the workers and safety belts, forming the second labeling result B.

[0096] This invention uses a target detection model trained on a Mask R-CNN network to detect enhanced channel images and obtain labeled results A and B, which can solve the problem of difficulty in identifying worker and safety belt labeled areas due to camera backlighting.

[0097] 6) The Otsu thresholding algorithm is used to perform thresholding segmentation on the images with labeled results A and B. The segmented binary images are labeled with connected components, and only the larger areas (i.e., areas greater than a set value) are selected. Small, redundant objects in the background are removed, and finally, the retained pixels are mapped back to the original image, thus achieving the segmentation of workers and safety belts in the image. Figure 3 As shown, the marked area for the workers is denoted as marked area C, and the marked area for the safety belt is denoted as marked area D.

[0098] The Otsu threshold segmentation algorithm of this invention can segment images and solve the problem of complex background interference caused by the diverse shapes and colors of substation switches, surge arresters and their supports.

[0099] 7) Through logical calculation, obtain the portion of marked region D after removing marked region C, thus obtaining the target marked region, and denoted as marked region E. See [link to relevant documentation]. Figure 4 The logical calculation formula for the pixels contained in the marked region E is as follows:

[0100]

[0101] Where x represents any pixel.

[0102] 8) Obtain the pixel height in the marked region C, i.e., the vertical coordinate y of the highest pixel. C Subtract the vertical coordinate y' of the lowest pixel C , denoted as h C The specific formula is as follows:

[0103] h C =y C -y' C

[0104] 9) Obtain the height y of the highest pixel in the marked region E. E Since the seatbelt interruption is always dangling, the highest pixel height y E This can be considered as an effective attachment point for the seat belt.

[0105] 10) The logic for determining whether a seatbelt is used when it is low-mounted and high-positioned is as follows:

[0106] y E <y’ C +0.75h C

[0107] That is, if the pixel height of the safety belt attachment point is lower than 75% of the pixel height of the worker's area, it is considered that the safety belt is used at a low position.

[0108] 11) When the safety belt is used at a lower position than it is attached, an alarm should be raised to the workers and managers.

[0109] Example 2

[0110] To implement the method corresponding to Embodiment 1 above and achieve the corresponding functions and technical effects, a detection system for low-attachment, high-use safety belts for high-altitude operations is provided below. For example... Figure 5 As shown, the system includes:

[0111] The image acquisition module 501 is used to acquire the image to be detected; the image to be detected is acquired by a camera located above the working plane of the target work site, facing the workers and safety belts.

[0112] The image enhancement module 502 is used to preprocess the image to be detected using an image enhancement algorithm to obtain the enhanced channel image to be tested.

[0113] The target detection module 503 is used to detect the image of the enhanced channel to be tested using a target detection model, and obtain a first marking result and a second marking result; the first marking result is the marking of the worker area and the marking of the safety belt area in the image of the enhanced channel to be tested; the second marking result is the marking of the worker boundary and the marking of the safety belt boundary in the image of the enhanced channel to be tested.

[0114] The image segmentation module 504 is used to segment the enhanced channel image to be tested, which has the first marking result and the second marking result, using an image segmentation algorithm to obtain the worker marking area and the safety belt marking area.

[0115] The area determination module 505 is used to remove the portion of the safety belt marking area that overlaps with the worker marking area to obtain the target marking area.

[0116] The height determination module 506 is used to determine a first height value and a second height value based on the worker-marked area, and to determine a third height value based on the target marked area; the first height value is the highest pixel height of the worker-marked area; the second height value is the lowest pixel height of the worker-marked area; and the third height value is the highest pixel height of the target marked area.

[0117] The safety detection module 507 is used to determine whether there is a phenomenon of low hanging and high use of safety belts at the target work site based on the first height value, the second height value and the third height value.

[0118] Example 3

[0119] This invention also provides an electronic device, including a memory and a processor. The memory stores a computer program, and the processor runs the computer program to enable the electronic device to perform the detection method for low-hanging, high-use safety belts in high-altitude operations as described in Embodiment 1. The electronic device may be a server.

[0120] In addition, the present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the detection method for low-hanging and high-use of safety belts for high-altitude operations in Embodiment 1.

[0121] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0122] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for detecting the low-hanging, high-use of safety belts for high-altitude operations, characterized in that, The method includes: Acquire the image to be detected; the image to be detected is acquired by a camera located above the working plane of the target work site, facing the workers and safety belts; An image enhancement algorithm is used to preprocess the image to be detected to obtain the enhanced channel image to be tested; An object detection model is used to detect the image of the enhancement channel to be tested, and a first labeling result and a second labeling result are obtained. The first labeling result is the worker area label and the safety belt area label in the image of the enhancement channel to be tested. The second labeling result is the worker boundary label and the safety belt boundary label in the image of the enhancement channel to be tested. The object detection model is trained based on the Mask R-CNN network. An image segmentation algorithm is used to segment the image of the enhanced channel to be tested, which contains the first marking result and the second marking result, to obtain the worker marking area and the safety belt marking area; Remove the portion of the safety belt marking area that overlaps with the worker marking area to obtain the target marking area; A first height value and a second height value are determined based on the worker-marked area, and a third height value is determined based on the target marked area; the first height value is the highest pixel height of the worker-marked area; the second height value is the lowest pixel height of the worker-marked area; and the third height value is the highest pixel height of the target marked area. Based on the first height value, the second height value, and the third height value, determine whether there is a phenomenon of low-hanging and high-use of safety belts at the target work site.

2. The detection method for low-hanging, high-use safety belts for high-altitude operations according to claim 1, characterized in that, The step of determining whether the target work site has a situation where the safety belt is used at a lower position than its intended purpose based on the first height value, the second height value, and the third height value specifically includes: The pixel height of the worker area is determined based on the first height value and the second height value; The pixel height of the seat belt attachment point is determined based on the third height value and the second height value; Determine whether the pixel height of the safety belt attachment point is less than a set multiple of the pixel height of the worker area, and obtain the determination result; the set multiple is 0.75 times. If the judgment result is yes, then it is determined that there is a phenomenon of low-hanging and high-use of safety belts at the target work site; If the judgment result is negative, then it is determined that there is no phenomenon of low-hanging and high-use of safety belts at the target work site.

3. The detection method for low-hanging, high-use safety belts for high-altitude operations according to claim 1, characterized in that, The step of employing an image enhancement algorithm to preprocess the image to be detected, thereby obtaining the image of the enhanced channel to be tested, specifically includes: The image to be detected is subjected to RGB channel separation, and the intensity channel image is determined based on the separated blue channel image, green channel image, and red channel image; A multi-scale retinal enhancement algorithm is used to enhance the intensity channel image to obtain an initial enhanced image; According to the set magnification factor, the pixel values ​​of each channel in the initial enhanced image are magnified to obtain the image of the enhanced channel to be tested.

4. The detection method for low-hanging, high-use safety belts for high-altitude operations according to claim 1, characterized in that, The target detection model includes a first target detection layer and a second target detection layer connected in series. The first target detection layer is used to detect the image of the enhanced channel to be tested and obtain a first labeling result; The second target detection layer is used to detect the image of the enhanced channel to be tested with the first labeling result to obtain the second labeling result.

5. The detection method for low-hanging, high-use safety belts for high-altitude operations according to claim 1, characterized in that, The image segmentation algorithm is used to segment the image of the enhancement channel to be tested, which contains the first and second labeling results, to obtain the worker labeling region and the safety belt labeling region, specifically including: The Otsu threshold segmentation algorithm is used to segment the image of the enhanced channel to be tested, which contains the first and second labeling results, to obtain the worker area and the safety belt area.

6. The detection method for low-hanging, high-use safety belts for high-altitude operations according to claim 2, characterized in that, The specific formula for determining the pixel height of the worker area based on the first height value and the second height value is as follows: h C =y C -y’ C Among them, h C The pixel height of the worker area, y C For the first height value, y' C This is the second altitude value.

7. The detection method for low-hanging, high-use safety belts for high-altitude operations according to claim 2, characterized in that, The specific formula for determining the pixel height of the seat belt attachment point based on the third height value and the second height value is as follows: and' E / and E -and' C Among them, y' E For the pixel height of the seatbelt attachment point, y E The third height value, y' C This is the second altitude value.

8. A detection system for low-hanging, high-use safety belts for high-altitude operations, characterized in that, The system includes: The image acquisition module is used to acquire the image to be detected; the image to be detected is acquired by a camera located above the working plane of the target work site, facing the workers and safety belts; The image enhancement module is used to preprocess the image to be detected using an image enhancement algorithm to obtain the enhanced channel image to be tested; The target detection module is used to detect the image of the enhanced channel to be tested using a target detection model, and obtain a first labeling result and a second labeling result; the first labeling result is the worker area label and the safety belt area label in the image of the enhanced channel to be tested; the second labeling result is the worker boundary label and the safety belt boundary label in the image of the enhanced channel to be tested. The image segmentation module is used to segment the image of the enhanced channel to be tested, which contains the first marking result and the second marking result, using an image segmentation algorithm to obtain the worker marking area and the safety belt marking area; The region determination module is used to remove the portion of the safety belt marking region that overlaps with the worker marking region to obtain the target marking region; A height determination module is used to determine a first height value and a second height value based on the worker-marked area, and to determine a third height value based on the target marked area; the first height value is the highest pixel height of the worker-marked area; the second height value is the lowest pixel height of the worker-marked area; and the third height value is the highest pixel height of the target marked area. The safety detection module is used to determine whether there is a phenomenon of low-hanging and high-use of safety belts at the target work site based on the first height value, the second height value, and the third height value.

9. An electronic device, characterized in that, The device includes a memory and a processor, the memory being used to store a computer program, and the processor running the computer program to cause the electronic device to perform the detection method for low-hanging, high-use safety belts for high-altitude operations as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores a computer program, which, when executed by a processor, implements the detection method for low-hanging, high-use safety belts for high-altitude operations as described in any one of claims 1 to 7.

Images