An edge protection position detection method, device, equipment and medium

Abstract

The application discloses a kind of edge protection position detection method, device, equipment and medium, by the edge protection detection model of well-trained, dangerous area and edge protection in monitoring video frame are identified;Respectively, the outline of the dangerous area and the edge protection is extracted, and the shortest distance between dangerous area outline and edge protection outline is determined;According to the shortest distance, whether the edge protection placement position is qualified is determined.It is realized to accurately separate and classify the edge protection and dangerous area in construction site, greatly improve the accuracy of safety monitoring, at the same time, the outline analysis and distance calculation introduced effectively evaluate whether the edge protection setting is up to standard, so as to comprehensively improve the safety management level of construction site.

Description

Technical Field

[0001] This application relates to the field of edge protection technology, specifically to a method, device, equipment, and medium for detecting the edge protection location. Background Technology

[0002] In construction scenarios, edge protection is a crucial measure to ensure the safety of workers and equipment. Currently, detection technologies for edge protection mainly include the following aspects: First, presence detection, such as using sensors or image recognition technology to detect whether there are guardrails around openings; second, edge protection status monitoring, such as real-time monitoring of the physical state of the guardrail itself, such as voltage fluctuations, to ensure that the guardrail is in normal working condition; and third, dangerous behavior identification, such as detecting whether there are dangerous actions such as people leaning against or climbing the guardrail.

[0003] However, existing technical solutions have certain limitations in practical applications. Current solutions typically assume that edge protection facilities are correctly installed around the danger zone as required. However, in actual construction, situations may arise where guardrails are installed but not placed around the danger zone, or the guardrails are too far from the edge of the danger zone, or even completely off-center. In such cases, although existing technologies perform presence checks on edge protection, their protective function cannot be fully realized because the guardrails are not in an effective protective position. Safety hazards remain and may even be overlooked, potentially leading to accidents. Summary of the Invention

[0004] This application provides a method, device, equipment, and medium for detecting the location of edge protection, which can effectively detect whether the location of edge protection meets the standards, thereby improving the safety management level of the construction site.

[0005] In a first aspect, embodiments of this application provide a method for detecting the location of edge protection, the method comprising:

[0006] By using a trained edge protection detection model, dangerous areas and edge protection can be identified in surveillance video frames.

[0007] Extract the outlines of the hazardous area and the edge protection respectively, and determine the shortest distance between the outline of the hazardous area and the outline of the edge protection;

[0008] The suitability of the edge protection placement position is determined based on the shortest distance.

[0009] In conjunction with the first aspect, in one implementation, a trained edge protection detection model is used to identify hazardous areas and edge protection in surveillance video frames, including:

[0010] By using a trained edge protection detection model, dangerous areas and edge protection in surveillance video frames are identified, and dangerous area masks and edge protection masks are obtained.

[0011] In conjunction with the first aspect, in one implementation, before identifying hazardous areas and edge protection in the surveillance video frame using the trained edge protection detection model, the method further includes:

[0012] Images of the construction site were captured from multiple perspectives, under different lighting conditions, and in different weather conditions.

[0013] The hazardous areas and edge protection in the construction site scene images are marked;

[0014] The edge protection detection model was trained using labeled construction site scene images to obtain a trained edge protection detection model.

[0015] In conjunction with the first aspect, in one implementation, the outlines of the hazardous area and the edge protection are extracted respectively, and the shortest distance between the hazardous area outline and the edge protection outline is determined, including:

[0016] Extract the hazardous area contour point set based on the hazardous area mask, and extract the edge protection contour point set based on the edge protection mask;

[0017] Calculate the Euclidean distance between each edge protection contour point in the edge protection contour point set and each danger area contour point in the danger area contour point set, and take the smallest value among all Euclidean distances as the shortest distance.

[0018] In conjunction with the first aspect, in one implementation, the step of calculating the Euclidean distance between each edge protection contour point in the edge protection contour point set and each danger zone contour point in the danger zone contour point set, and taking the smallest value among all Euclidean distances as the shortest distance, includes:

[0019] Choose either the edge protection contour point set or the hazardous area contour point set as the target contour point set;

[0020] Construct a KD tree based on the target contour point set;

[0021] For each contour point in the non-target contour point set, the KD tree is used to perform a nearest neighbor search to obtain the Euclidean distance from the contour point to the nearest neighbor in the target contour point set.

[0022] Iterate through all non-target contour points and their corresponding nearest neighbors using Euclidean distances, and take the smallest Euclidean distance as the shortest distance.

[0023] In conjunction with the first aspect, in one implementation, determining whether the edge protection placement position is qualified based on the shortest distance includes:

[0024] If the shortest distance is greater than the distance threshold, then the placement of the edge protection is determined to be unqualified.

[0025] If the shortest distance is less than or equal to a preset distance threshold, then the placement of the edge protection is deemed qualified.

[0026] In conjunction with the first aspect, in one implementation, the method further includes:

[0027] After determining that the edge protection placement is unqualified, a warning signal is issued indicating that the edge protection is placed too far away.

[0028] Secondly, embodiments of this application provide an edge protection position detection device, the edge protection position detection device comprising:

[0029] The identification module is used to identify dangerous areas and edge protection in surveillance video frames using a trained edge protection detection model.

[0030] The first determining module is used to extract the outlines of the hazardous area and the edge protection respectively, and to determine the shortest distance between the outline of the hazardous area and the outline of the edge protection.

[0031] The second determining module is used to determine whether the placement position of the edge protection is qualified based on the shortest distance.

[0032] Thirdly, embodiments of this application provide an edge protection position detection device, characterized in that the edge protection position detection device includes a processor, a memory, and an edge protection position detection program stored in the memory and executable by the processor, wherein when the edge protection position detection program is executed by the processor, it implements the steps of the edge protection position detection method as described in any of the preceding claims.

[0033] Fourthly, embodiments of this application provide a computer-readable storage medium, characterized in that the computer-readable storage medium stores an edge protection position detection program, wherein when the edge protection position detection program is executed by a processor, it implements the steps of the edge protection position detection method as described in any of the preceding claims.

[0034] This application provides a method, apparatus, equipment, and medium for detecting the location of edge protection. Using a trained edge protection detection model, it identifies hazardous areas and edge protection in monitoring video frames; extracts the contours of the hazardous areas and the edge protection respectively; and determines the shortest distance between the contours of the hazardous areas and the edge protection; based on the shortest distance, it determines whether the placement of the edge protection is qualified. This achieves precise segmentation and classification of edge protection and hazardous areas in construction sites, greatly improving the accuracy of safety monitoring. Simultaneously, the introduced contour analysis and distance calculation effectively assess whether the edge protection settings meet standards, thereby comprehensively improving the safety management level of construction sites. Attached Figure Description

[0035] Figure 1 This is a flowchart illustrating an embodiment of the edge protection location detection method of this application;

[0036] Figure 2 This is a schematic flowchart of another embodiment of the edge protection location detection method of this application;

[0037] Figure 3 This is a functional module diagram of an embodiment of the edge protection position detection device of this application;

[0038] Figure 4 This is a schematic diagram of the hardware structure of the edge protection position detection device involved in the embodiment of this application. Detailed Implementation

[0039] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0040] First, some of the technical terms used in this application will be explained to help those skilled in the art understand this application.

[0041] KD-tree (K-dimensional tree): A tree-like data structure that stores instance points in k-dimensional space for fast retrieval. It is mainly used for searching key data in multi-dimensional space.

[0042] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0043] In a first aspect, embodiments of this application provide a method for detecting the location of edge protection.

[0044] In one embodiment, reference is made to Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the edge protection location detection method of this application. Figure 1 As shown, the edge protection location detection method includes:

[0045] Step S101: Identify dangerous areas and edge protection in the monitoring video frames using the trained edge protection detection model.

[0046] It is worth noting that the types of hazardous areas and edge protection in this embodiment can be set according to actual needs. For example, hazardous areas may include trenches, foundation pits, and openings in construction sites, as well as areas with electrical equipment and road construction areas. Edge protection may include guardrails, protective covers, safety warning signs, and other protective facilities.

[0047] In one embodiment, before step S101, the method further includes: acquiring construction site scene images from multiple perspectives under different lighting and weather conditions; marking dangerous areas and edge protection in the construction site scene images; and using the marked construction site scene images to train the edge protection detection model to obtain a trained edge protection detection model.

[0048] Exemplary, such as Figure 2 As shown, on the training side, multiple construction site scene images covering different lighting and weather conditions can be obtained through oblique photography by drones and multi-angle shooting by ground cameras. These images include the hazardous areas to be identified and the types of edge protection. Data annotation tools, such as Labelme, are used to label the various types of hazardous areas and edge protection in the captured construction site scene images, thus processing the training data. Next, the labeled construction site scene images are divided into training and testing sets to construct the dataset.

[0049] In this embodiment, the edge protection detection model can be trained using an instance segmentation algorithm. That is, the kernel of the edge protection detection model includes an instance segmentation model, such as YOLOv12. The training set is input into the instance segmentation model for training. During training, the instance segmentation model learns how to identify hazardous scenes and edge protection from the input image and generates corresponding segmentation masks. The parameters of the instance segmentation model are continuously adjusted during training to minimize the error between the hazardous scenes and edge protection outputs of the instance segmentation model and the actual labeled hazardous scenes and edge protection. The instance segmentation model is then tested using a test set until the error between the instance segmentation model output and the actual labeled data is within a preset error range. This completes the training of the instance segmentation model, resulting in a trained edge protection detection model capable of accurately segmenting and classifying hazardous areas and edge protection.

[0050] Optionally, the training of the edge protection detection model can be completed offline on the training side, while edge protection location detection is performed online on the inference side. During edge protection location detection, on-site image data such as monitoring video frames can be directly input into the trained edge protection detection model. The model identifies hazardous areas and edge protections in the monitoring video frames, obtaining hazardous area masks and edge protection masks. Then, contour analysis and distance calculations are performed based on the hazardous area masks and edge protection masks to ultimately determine whether the edge protection placement is qualified, thus outputting an early warning result. The hazardous area mask and edge protection mask can be pixel-level masks.

[0051] Step S102: Extract the outlines of the hazardous area and the edge protection respectively, and determine the shortest distance between the outline of the hazardous area and the outline of the edge protection.

[0052] In one embodiment, step S102 specifically includes: extracting a set of dangerous area contour points based on the dangerous area mask, extracting a set of edge protection contour points based on the edge protection mask; calculating the Euclidean distance between each edge protection contour point in the edge protection contour point set and each dangerous area contour point in the dangerous area contour point set, and taking the smallest value among all Euclidean distances as the shortest distance.

[0053] Exemplary, such as Figure 2 As shown, OpenCV can be used for contour analysis to extract the hazard area contour from the hazard area mask and the edge protection contour from the edge protection mask. Then, `cv2.findContours` is used to generate a hazard area contour point set based on the hazard area contour and an edge protection contour point set based on the edge protection contour. For each hazard area and edge protection in the surveillance video frame, the corresponding contour point set can be represented as... Where P is the set of contour points, and each point in the set is a contour point. Represents the position coordinates of a point on the contour, such as , For contour points x-coordinate For contour points The ordinate is the vertical axis. The coordinate system has the top left corner of the image as the origin (0,0), with the x-axis pointing to the right and the y-axis pointing downwards.

[0054] Furthermore, the step of calculating the Euclidean distance between each edge protection contour point in the edge protection contour point set and each danger zone contour point in the danger zone contour point set, and taking the smallest Euclidean distance as the shortest distance, includes: selecting either the edge protection contour point set or the danger zone contour point set as the target contour point set; constructing a KD tree based on the target contour point set; for each contour point not in the target contour point set, performing a nearest neighbor search using the KD tree to obtain the Euclidean distance from the contour point to the nearest neighbor in the target contour point set; traversing all the Euclidean distances between the non-target contour points and their corresponding nearest neighbors, and taking the smallest Euclidean distance as the shortest distance.

[0055] As an example, when performing distance calculations, if a trench contour point set is selected as the target contour point set, the remaining contour point set is the non-target contour point set. A KD tree is constructed based on the trench contour point set Pg. For the edge protection contour point set... Each contour point Use a KD-tree to query the set of points corresponding to the groove profile. nearest neighbor The nearest neighbor can be found using the nearest neighbor search function of the KD tree, which returns the nearest neighbor by Euclidean distance and its corresponding Euclidean distance.

[0056] Given two points and The Euclidean distance D between them is calculated using the following formula:

[0057]

[0058] Where D is and Euclidean distance, for x-coordinate for x-coordinate for The ordinate, for The ordinate.

[0059] Iterate through all edge protection contour points in the edge protection contour point set, and calculate the Euclidean distance from each edge protection contour point to its nearest neighbor in the trench contour point set using the method described above. The minimum value of all calculated Euclidean distances is taken as the shortest distance between the edge protection and the hazardous area (trench).

[0060]

[0061] in, M represents the shortest distance, and M is the number of edge protection contour points. It represents the Euclidean distance from the i-th edge protection contour point to the nearest neighbor point in the trench contour point set.

[0062] Using KD-trees for nearest neighbor search can improve the efficiency and speed of searching for nearest neighbors, thus allowing for a faster determination of the shortest distance between the hazardous area and the edge protection.

[0063] Step S103: Determine whether the placement of the edge protection is qualified based on the shortest distance.

[0064] Specifically, if the shortest distance Greater than the distance threshold ,Right now If the shortest distance is less than or equal to a preset distance threshold, then the placement of the edge protection is deemed unqualified; If the condition is met, the placement of the edge protection is deemed acceptable. The distance threshold can be set according to actual safety requirements.

[0065] Furthermore, after determining that the edge protection placement is unqualified, a warning signal is issued indicating that the edge protection is placed too far away. This warning signal includes: an on-site audible alarm to alert personnel to the danger zone, and sending a warning message to management personnel via SMS or email to instruct them to adjust the edge protection's position, placing it at a suitable distance from the danger zone. Edge protection placed too far from the danger zone will fail to provide adequate safety protection, thereby improving the safety management level of the construction site.

[0066] In one optional embodiment, when the edge protection detection model identifies a dangerous area in a monitoring video frame but no edge protection exists, it issues a warning signal indicating that there is no edge protection near the dangerous area.

[0067] The edge protection location detection method provided in this application utilizes deep learning technology to accurately segment and classify edge protection and hazardous area-related targets at construction sites, significantly improving the accuracy of safety monitoring. Furthermore, this method incorporates contour analysis technology and combines it with KD-tree calculations for minimum distance, enabling rapid and effective assessment of whether edge protection measures meet safety standards and timely issuance of hazard warnings, thus comprehensively enhancing the safety management level of construction sites.

[0068] Secondly, embodiments of this application also provide an edge protection position detection device.

[0069] In one embodiment, reference is made to Figure 3 , Figure 3 This is a functional module diagram of an embodiment of the edge protection position detection device of this application. Figure 3 As shown, the edge protection position detection device includes:

[0070] The identification module is used to identify dangerous areas and edge protection in surveillance video frames using a trained edge protection detection model.

[0071] The first determining module is used to extract the outlines of the hazardous area and the edge protection respectively, and to determine the shortest distance between the outline of the hazardous area and the outline of the edge protection.

[0072] The second determining module is used to determine whether the placement position of the edge protection is qualified based on the shortest distance.

[0073] Furthermore, in one embodiment, the identification module is also used for:

[0074] By using a trained edge protection detection model, dangerous areas and edge protection in surveillance video frames are identified, and dangerous area masks and edge protection masks are obtained.

[0075] Furthermore, in one embodiment, the device further includes a training module, which is used for:

[0076] Images of the construction site were captured from multiple perspectives, under different lighting conditions, and in different weather conditions.

[0077] The hazardous areas and edge protection in the construction site scene images are marked;

[0078] The edge protection detection model was trained using labeled construction site scene images to obtain a trained edge protection detection model.

[0079] Furthermore, in one embodiment, the first determining module is further configured to:

[0080] Extract the hazardous area contour point set based on the hazardous area mask, and extract the edge protection contour point set based on the edge protection mask;

[0081] Calculate the Euclidean distance between each edge protection contour point in the edge protection contour point set and each danger area contour point in the danger area contour point set, and take the smallest value among all Euclidean distances as the shortest distance.

[0082] Furthermore, in one embodiment, the first determining module is further configured to:

[0083] Choose either the edge protection contour point set or the hazardous area contour point set as the target contour point set;

[0084] Construct a KD tree based on the target contour point set;

[0085] For each contour point in the non-target contour point set, the KD tree is used to perform a nearest neighbor search to obtain the Euclidean distance from the contour point to the nearest neighbor in the target contour point set.

[0086] Iterate through all non-target contour points and their corresponding nearest neighbors using Euclidean distances, and take the smallest Euclidean distance as the shortest distance.

[0087] Furthermore, in one embodiment, the second determining module is further configured to:

[0088] If the shortest distance is greater than the distance threshold, then the placement of the edge protection is determined to be unqualified.

[0089] If the shortest distance is less than or equal to a preset distance threshold, then the placement of the edge protection is deemed qualified.

[0090] Furthermore, in one embodiment, the edge protection position detection device further includes:

[0091] The early warning module is used to issue an early warning signal indicating that the edge protection is placed too far away after determining that the edge protection placement position is unqualified.

[0092] The functions of each module in the aforementioned edge protection position detection device correspond to the steps in the aforementioned edge protection position detection method embodiment, and their functions and implementation processes will not be described in detail here.

[0093] Thirdly, embodiments of this application provide an edge protection position detection device, which can be a personal computer (PC), laptop computer, server, or other device with data processing capabilities.

[0094] Reference Figure 4, Figure 4 This is a schematic diagram of the hardware structure of the edge protection position detection device involved in the embodiments of this application. In the embodiments of this application, the edge protection position detection device may include a processor, a memory, a communication interface, and a communication bus.

[0095] The communication bus can be of any type and is used to interconnect the processor, memory, and communication interface.

[0096] The communication interface includes input / output (I / O) interfaces, physical interfaces, and logical interfaces used for interconnecting internal components of the edge protection position detection equipment, as well as interfaces used for interconnecting the edge protection position detection equipment with other devices (such as other computing devices or user equipment). Physical interfaces can be Ethernet interfaces, fiber optic interfaces, ATM interfaces, etc.; user equipment can be displays, keyboards, etc.

[0097] Memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0098] The processor can be a general-purpose processor, which can call the edge protection position detection program stored in the memory and execute the edge protection position detection method provided in the embodiments of this application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed when the edge protection position detection program is called can be referred to in the various embodiments of the edge protection position detection method of this application, and will not be repeated here.

[0099] Those skilled in the art will understand that Figure 4 The hardware structure shown does not constitute a limitation of this application and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0100] Fourthly, embodiments of this application also provide a computer-readable storage medium.

[0101] The present application provides a computer-readable storage medium storing an edge protection position detection program, wherein when the edge protection position detection program is executed by a processor, it implements the steps of the edge protection position detection method described above.

[0102] The method implemented when the edge protection position detection procedure is executed can be referred to in various embodiments of the edge protection position detection method of this application, and will not be repeated here.

[0103] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0104] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.

[0105] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.

[0106] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.

[0107] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.

[0108] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods described in the various embodiments of this application.

[0109] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for detecting the location of edge protection, characterized in that, The method for detecting the edge protection location includes: By using a trained edge protection detection model, dangerous areas and edge protection can be identified in surveillance video frames. Extract the outlines of the hazardous area and the edge protection respectively, and determine the shortest distance between the outline of the hazardous area and the outline of the edge protection; Determine whether the placement of the edge protection is qualified based on the shortest distance; Among these features, the trained edge protection detection model identifies hazardous areas and edge protection in surveillance video frames, including: By using a trained edge protection detection model, dangerous areas and edge protection in surveillance video frames are identified, and dangerous area masks and edge protection masks are obtained. This includes extracting the contours of the hazardous area and the edge protection, and determining the shortest distance between the hazardous area contour and the edge protection contour, including: Extract the hazardous area contour point set based on the hazardous area mask, and extract the edge protection contour point set based on the edge protection mask; Calculate the Euclidean distance between each edge protection contour point in the edge protection contour point set and each danger area contour point in the danger area contour point set, and take the smallest value among all Euclidean distances as the shortest distance.

2. The edge protection position detection method as described in claim 1, characterized in that, Before identifying hazardous areas and edge protection in surveillance video frames using the trained edge protection detection model, the process also includes: Images of the construction site were captured from multiple perspectives, under different lighting conditions, and in different weather conditions. The hazardous areas and edge protection in the construction site scene images are marked; The edge protection detection model was trained using labeled construction site scene images to obtain a trained edge protection detection model.

3. The edge protection position detection method as described in claim 1, characterized in that, The calculation of the Euclidean distance between each edge protection contour point in the edge protection contour point set and each hazardous area contour point in the hazardous area contour point set, and the smallest of all Euclidean distances as the shortest distance, includes: Choose either the edge protection contour point set or the hazardous area contour point set as the target contour point set; Construct a KD tree based on the target contour point set; For each contour point in the non-target contour point set, the KD tree is used to perform a nearest neighbor search to obtain the Euclidean distance from the contour point to the nearest neighbor in the target contour point set. Iterate through all non-target contour points and their corresponding nearest neighbors using Euclidean distances, and take the smallest Euclidean distance as the shortest distance.

4. The edge protection position detection method as described in claim 1, characterized in that, Determining whether the placement of the edge protection is qualified based on the shortest distance includes: If the shortest distance is greater than the distance threshold, then the placement of the edge protection is determined to be unqualified. If the shortest distance is less than or equal to a preset distance threshold, then the placement of the edge protection is deemed qualified.

5. The edge protection position detection method as described in claim 1, characterized in that, The method also includes: After determining that the edge protection placement is unqualified, a warning signal is issued indicating that the edge protection is placed too far away.

6. A device for detecting the position of edge protection, characterized in that, The edge protection position detection device includes: The identification module is used to identify dangerous areas and edge protection in surveillance video frames using a trained edge protection detection model. The first determining module is used to extract the outlines of the hazardous area and the edge protection respectively, and to determine the shortest distance between the outline of the hazardous area and the outline of the edge protection. The second determining module is used to determine whether the placement position of the edge protection is qualified based on the shortest distance; The identification module is further used for: By using a trained edge protection detection model, dangerous areas and edge protection in surveillance video frames are identified, and dangerous area masks and edge protection masks are obtained. The first determining module is further configured to: Extract the hazardous area contour point set based on the hazardous area mask, and extract the edge protection contour point set based on the edge protection mask; Calculate the Euclidean distance between each edge protection contour point in the edge protection contour point set and each danger area contour point in the danger area contour point set, and take the smallest value among all Euclidean distances as the shortest distance.

7. An edge protection position detection device, characterized in that, The edge protection position detection device includes a processor, a memory, and an edge protection position detection program stored in the memory and executable by the processor, wherein when the edge protection position detection program is executed by the processor, it implements the steps of the edge protection position detection method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores an edge protection position detection program, wherein when the edge protection position detection program is executed by a processor, it implements the steps of the edge protection position detection method as described in any one of claims 1 to 5.

Images