Stringing construction safety monitoring method and system based on image processing

Abstract

The invention relates to the technical field of electric power and information processing, and provides a stringing construction safety monitoring method and system based on image processing, and the method comprises the steps: obtaining a top view image obtained through the shooting of a first target stringing construction region by an unmanned plane; whether an intersection exists between a target frame of a constructor operating the tail rope of the maneuvering winch and a target frame of the rope loop is judged based on the top view image; if yes, calculating the area of the intersection region; if the area of the intersection area is larger than a preset area threshold value, the area of the rope loop is calculated; calculating the area ratio of the intersection area to the rope loop; identifying the operation state of the motorized winching machine; determining a safety risk level according to the operation state of the motorized winching machine and the area ratio of the intersection area to the rope loop; determining a warning strategy according to the security risk level; and through a voice playing module of the unmanned aerial vehicle, a constructor operating the tail rope of the maneuvering winching machine is warned according to the warning strategy. According to the application, the use safety of the motorized winching machine can be automatically monitored and warned quickly and accurately.

Description

Technical Field

[0001] The present application relates to the field of electric power and information processing technology, and in particular to a method and system for monitoring the safety of wiring construction based on image processing. Background Art

[0002] In the power industry, line construction refers to the installation process of installing overhead lines with connecting fittings and insulator strings on assembled poles and towers. Line construction is usually carried out at high altitudes and in complex environments, and there are many potential safety hazards, such as falling from heights, being hit by objects, and electric shock. By conducting safety monitoring of line construction, unsafe factors at the construction site can be discovered in a timely manner, such as illegal operations by personnel, so as to effectively prevent accidents and protect the lives of construction workers. In order to effectively prevent safety accidents, intelligent monitoring systems with seat belt monitoring and / or safety helmet monitoring functions are currently available on the market. However, with the growing demand for safety monitoring in various scenarios, the functions of existing intelligent monitoring systems can no longer fully meet user needs. In power line construction, a motorized winch is an indispensable equipment. Whether it is the construction of high-voltage transmission lines or low-voltage distribution lines, it is necessary to use it to pull and tighten cables. At present, the safety of the use of motorized capstans is monitored manually. Specifically, manual inspection is performed to check whether the construction personnel operating the tail rope of the motorized capstan is in the rope loop of the tail rope. If the construction personnel is in the rope loop of the tail rope, the construction personnel is ordered to leave the rope loop and stand with both feet outside the rope loop. However, the manual method has the problem of low efficiency. Therefore, how to realize automated monitoring of the safety of the use of motorized capstans, especially to realize fast and accurate monitoring of whether the construction personnel operating the tail rope of the motorized capstan is in the rope loop of the tail rope, is a technical problem that needs to be solved urgently. Summary of the invention

[0003] In view of the above technical problems, the purpose of this application is to provide a wireline construction safety monitoring method and system based on image processing, aiming to quickly and accurately monitor whether the construction personnel operating the tail rope of the motorized capstan are located within the tail rope loop.

[0004] In order to solve the above technical problems, the embodiment of the present application provides a wireline construction safety monitoring method based on image processing, which is applied to a drone and includes:

[0005] Acquire a bird's-eye view image of the first target cable construction area photographed by the drone; wherein the bird's-eye view image includes a motorized capstan, a tail rope of the motorized capstan, and construction personnel;

[0006] Using a bird's-eye view image of the first target wiring construction area photographed by the drone as a first image;

[0007] Based on the first image, the target detection model and the distance calculation method, the construction worker operating the tail rope of the motorized capstan is identified and the position information of the target frame of the construction worker operating the tail rope of the motorized capstan is output; wherein the target frame output by the target detection model is a rectangular frame, the image input by the target detection model is a rectangular image, and the four sides of the target frame are respectively parallel to the four sides of the input image;

[0008] Identify the rope loop of the tail rope of the motorized capstan based on the first image and the target detection model and output the position information of the target frame of the rope loop;

[0009] Determining whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop intersect according to the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop;

[0010] If so, calculate the area of ​​the intersection;

[0011] If the area of ​​the intersection region is greater than a preset area threshold, then calculating the area of ​​the rope loop;

[0012] Dividing the area of ​​the intersection region by the area of ​​the rope loop to obtain a ratio of the area of ​​the intersection region to the area of ​​the rope loop;

[0013] Identifying the operating status of the motorized capstan;

[0014] Determining the safety risk level according to the operating state of the motorized capstan and the ratio of the intersection area to the area of ​​the rope loop;

[0015] Determining a warning strategy based on the security risk level;

[0016] The construction personnel operating the tail rope of the motorized capstan are warned according to the warning strategy through the voice playback module of the drone; wherein the warning is used to warn the construction personnel operating the tail rope of the motorized capstan not to stand in the rope loop.

[0017] Further, the position information of the target frame of the construction worker operating the tail rope of the motorized capstan includes the lower left corner coordinates (x1, y1) of the target frame of the construction worker operating the tail rope of the motorized capstan and the upper right corner coordinates (x2, y2) of the target frame of the construction worker operating the tail rope of the motorized capstan, and the position information of the target frame of the rope loop includes the lower left corner coordinates (x3, y3) of the target frame of the rope loop and the upper right corner coordinates (x4, y4) of the target frame of the rope loop;

[0018] The step of judging whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop intersects according to the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop comprises:

[0019] Determine whether max(x1,x3)<=min(x2,x4) and max(y1,y3)<=min(y2,y4) are satisfied; where min means taking the minimum value and max means taking the maximum value;

[0020] If so, it is determined that the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have an intersection; otherwise, it is determined that the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have no intersection.

[0021] Furthermore, the step of calculating the area of ​​the intersection region includes:

[0022] The area of ​​the intersection region is calculated according to the following formula;

[0023] S=[min(x2,x4)-max(x1,x3)]×[min(y2,y4)-max(y1,y3)];

[0024] Where S is the area of ​​the intersection region.

[0025] In one embodiment, the step of identifying the loop of the tail rope of the motorized capstan based on the first image and the target detection model and outputting the position information of the target frame of the loop includes:

[0026] Identify the tail rope of the motorized capstan and the construction personnel through the target detection model, and output the position information of the tail rope of the motorized capstan and the position information of the construction personnel;

[0027] Calculate the shortest distance from the construction worker to the tail rope of the motorized capstan according to the position information of the tail rope of the motorized capstan and the position information of the construction worker;

[0028] Determining whether the shortest distance is less than a preset first threshold;

[0029] If it is less than, it is determined that the identified construction worker is a construction worker operating the tail rope of the motorized capstan. Otherwise, it is determined that the identified construction worker is not a construction worker operating the tail rope of the motorized capstan.

[0030] Further, the location information of the construction worker includes the center point coordinates of the construction worker target frame, the location information of the motorized capstan tail rope includes the location information of the motorized capstan tail rope target frame, and the step of calculating the shortest distance from the construction worker to the motorized capstan tail rope according to the location information of the motorized capstan tail rope and the location information of the construction worker includes:

[0031] Positioning the target frame of the motorized capstan tail rope according to the position information of the target frame of the motorized capstan tail rope;

[0032] Extracting the contour of the tail rope of the motorized capstan based on the target frame of the tail rope of the motorized capstan;

[0033] Calculate the distance between the center point coordinate of the construction worker target frame and each pixel point in the tail rope contour of the motorized capstan according to the center point coordinate of the construction worker target frame and the coordinate of each pixel point in the tail rope contour of the motorized capstan;

[0034] All the calculated distances are compared to obtain the shortest distance from the construction worker to the tail rope of the motorized capstan.

[0035] Furthermore, before the tail rope of the motorized capstan and the construction personnel are identified by the target detection model, the method further includes:

[0036] A large number of overhead images of the wiring construction area are taken by a drone; wherein the overhead image of the wiring construction area includes construction personnel, a motorized capstan, a tail rope of the motorized capstan, and a rope loop of the tail rope of the motorized capstan; wherein the overhead image is a rectangular image;

[0037] Classify and label each target object in the overhead image of the cable construction area to obtain a data set; wherein, classify and label each target object in the overhead image of the cable construction area includes classifying and labeling construction personnel, motorized capstan, motorized capstan tail rope, and rope loop of the motorized capstan tail rope; wherein, when labeling, the target frame is a rectangular frame, the target frame needs to frame the outline of each target object, and the four sides of the target frame are respectively parallel to the four sides of the overhead image;

[0038] The YOLO model is trained using the data set, and the training is stopped when the loss function converges to obtain the target detection model; wherein the target detection model is used to identify construction personnel, motorized capstans, motorized capstan tail ropes, and rope loops of motorized capstan tail ropes.

[0039] Further, after the step of judging whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have an intersection based on the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop, the method further includes:

[0040] If the target frame of the construction worker operating the tail rope of the motorized capstan does not intersect with the target frame of the rope loop, then counting the number of identified construction workers operating the tail rope of the motorized capstan;

[0041] Determine whether the number of construction workers operating the tail rope of the motorized capstan is greater than 2;

[0042] If it is not greater than, a second voice is played through the voice playback module on the drone to warn the construction worker operating the tail rope of the motorized capstan that he has violated the rules and that it is necessary to ensure that there are more than two construction workers operating the tail rope of the motorized capstan.

[0043] Furthermore, after the step of determining whether the number of construction workers operating the tail rope of the motorized capstan is greater than 2, the method further includes:

[0044] If there are more than 2 persons, the motorized winch is identified based on the target detection model;

[0045] Calculating the distance between the motorized capstan and each construction worker operating the tail rope of the motorized capstan, and comparing them to obtain the distance between the construction worker closest to the motorized capstan and the motorized capstan;

[0046] Determining whether the distance between the construction worker closest to the motorized capstan and the motorized capstan is greater than a preset safety distance;

[0047] If it is less than, a third voice is played through the voice playback module on the drone to warn the construction worker operating the tail rope of the motorized capstan that he has violated the rules and to ensure that the distance between the construction worker closest to the motorized capstan and the motorized capstan is greater than the safety distance.

[0048] The present application also provides a wireline construction safety monitoring system based on image processing, which is applied to a drone and includes:

[0049] An image acquisition module is used to acquire a bird's-eye view image of the first target cable construction area photographed by the drone; wherein the bird's-eye view image includes a motorized capstan, a tail rope of the motorized capstan, and construction personnel;

[0050] An equivalent module, configured to use the overhead image of the first target wiring construction area photographed by the drone as the first image;

[0051] A first recognition module is used to identify the construction worker who operates the tail rope of the motorized capstan based on the first image, the target detection model and the distance calculation method, and output the position information of the target frame of the construction worker who operates the tail rope of the motorized capstan; wherein the target frame output by the target detection model is a rectangular frame, the image input by the target detection model is a rectangular image, and the four sides of the target frame are respectively parallel to the four sides of the input image;

[0052] A second recognition module, which recognizes the rope loop of the tail rope of the motorized capstan based on the first image and the target detection model and outputs the position information of the target frame of the rope loop;

[0053] A first judgment module is used to judge whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have an intersection according to the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop;

[0054] A first calculation module is used to calculate the area of ​​the intersection region, if any;

[0055] A second calculation module, configured to calculate the area of ​​the rope loop if the area of ​​the intersection region is greater than a preset area threshold;

[0056] A third calculation module, configured to obtain a ratio of the area of ​​the intersection area to the area of ​​the rope loop by dividing the area of ​​the intersection area by the area of ​​the rope loop;

[0057] A third identification module, used for identifying the operating state of the motorized winch;

[0058] A safety risk level determination module, used to determine the safety risk level according to the operating state of the motorized capstan and the ratio of the intersection area to the area of ​​the rope loop;

[0059] A warning strategy determination module, used to determine a warning strategy according to the security risk level;

[0060] The warning module is used to warn the construction personnel who operate the tail rope of the motorized capstan according to the warning strategy through the voice playback module of the drone; wherein the warning is used to warn the construction personnel who operate the tail rope of the motorized capstan not to stand in the rope loop.

[0061] Further, the position information of the target frame of the construction worker operating the tail rope of the motorized capstan includes the lower left corner coordinates (x1, y1) of the target frame of the construction worker operating the tail rope of the motorized capstan and the upper right corner coordinates (x2, y2) of the target frame of the construction worker operating the tail rope of the motorized capstan, and the position information of the target frame of the rope loop includes the lower left corner coordinates (x3, y3) of the target frame of the rope loop and the upper right corner coordinates (x4, y4) of the target frame of the rope loop;

[0062] The step of judging whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop intersects according to the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop comprises:

[0063] Determine whether max(x1,x3)<=min(x2,x4) and max(y1,y3)<=min(y2,y4) are satisfied; where min means taking the minimum value and max means taking the maximum value;

[0064] If so, it is determined that the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have an intersection; otherwise, it is determined that the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have no intersection.

[0065] The image obtained from the top view angle of the present application can more intuitively show the relative position relationship between the construction workers and the rope loop. Since the rope loop of the tail rope is usually on the ground, the top view image can more easily distinguish whether the construction workers are inside or outside the rope loop. This position relationship is clear at a glance from the top view angle of the plane. Compared with observing from the side or other angles, it reduces the judgment error caused by occlusion, etc., and improves the accuracy of detection. In addition, by comparing whether the target frame of the construction worker and the target frame of the rope loop have an intersection, the position relationship between the construction worker and the rope loop in space can be intuitively understood. This judgment method based on the intersection of the target frame simplifies the complex three-dimensional spatial position relationship into a geometric relationship judgment on a two-dimensional plane, reduces the difficulty and complexity of the judgment, and improves the speed of detection. When the target frame has an intersection, it means that the construction worker and the rope loop have overlapping parts on the image plane, and it is possible that the construction worker is located in the rope loop. On the basis of judging that there is an intersection, the area of ​​the intersection area is further calculated and compared with the preset area threshold, so that the judgment is more accurate. If the intersection area is small, it cannot be determined that the construction worker is inside the rope loop, for example, if the shoulder partially overlaps the rope loop; only when the intersection area is larger than the preset area threshold can it be more reliably determined that the construction worker is inside the rope loop, avoiding misjudgment caused by slight overlap of the target frame. In addition, by calculating the ratio of the intersection area to the rope loop area, it is possible to more accurately determine the proportion of the construction worker in the rope loop. Combined with the operating status of the motorized capstan, it is possible to more accurately determine the safety risk level of the construction worker. Different warning strategies are used according to different safety risk levels, which is more targeted and greatly protects the life safety and physical health of the construction workers. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the technical solution of the present application, the drawings required for use in the implementation manner will be briefly introduced below. Obviously, the drawings described below are only some implementation manners of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0067] Figure 1 It is a flow chart of a wireline construction safety monitoring method based on image processing provided in an embodiment of the present application. DETAILED DESCRIPTION

[0068] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0069] It will be understood by those skilled in the art that, unless expressly stated, the singular forms "a", "an", "above", and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present application refers to the presence of features, integers, steps, operations, elements, modules, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, modules, components, and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be an intermediate element. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and / or" used herein includes all or any module and all combinations of one or more associated listed items.

[0070] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art to which this application belongs. It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with those in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless specifically defined as here.

[0071] See also Figure 1 The embodiment of the present application provides a wireline construction safety monitoring method based on image processing, which is applied to a drone and includes steps S1-S12.

[0072] S1. Obtaining a bird's-eye view image of a first target cable construction area photographed by a drone; wherein the bird's-eye view image includes a motorized capstan, a tail rope of the motorized capstan, and construction personnel;

[0073] S2, taking a bird's-eye view image of the first target cable construction area photographed by the drone as a first image;

[0074] S3, based on the first image, the target detection model and the distance calculation method, identifying the construction worker who operates the tail rope of the motorized capstan and outputting the position information of the target frame of the construction worker who operates the tail rope of the motorized capstan; wherein the target frame output by the target detection model is a rectangular frame, the image input by the target detection model is a rectangular image, and the four sides of the target frame are respectively parallel to the four sides of the input image;

[0075] S4, identifying the rope loop of the tail rope of the motorized capstan based on the first image and the target detection model and outputting the position information of the target frame of the rope loop;

[0076] S5, judging whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have an intersection according to the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop;

[0077] S6. If yes, calculate the area of ​​the intersection;

[0078] S7. If the area of ​​the intersection region is greater than a preset area threshold, then calculating the area of ​​the rope loop;

[0079] S8, dividing the area of ​​the intersection region by the area of ​​the rope loop to obtain a ratio of the area of ​​the intersection region to the area of ​​the rope loop;

[0080] S9, identifying the operating status of the motorized capstan;

[0081] S10, determining a safety risk level according to the operating state of the motorized capstan and the ratio of the intersection area to the area of ​​the rope loop;

[0082] S11, determining a warning strategy according to the security risk level;

[0083] S12. Warning the construction worker who operates the tail rope of the motorized capstan according to the warning strategy through the voice playback module of the drone; wherein the warning is used to warn the construction worker who operates the tail rope of the motorized capstan not to stand in the rope loop.

[0084] In step S1, the wiring construction mainly refers to the installation of wires, lightning arresters and other cables in a specific area according to the design requirements through a series of operations to a predetermined position, thereby establishing a line channel for power transmission or communication signal transmission. In the construction of power wiring, the motorized capstan is an indispensable equipment. Whether it is the construction of high-voltage transmission lines or low-voltage distribution lines, it is necessary to use it to pull and tighten the cables. The tail rope of the motorized capstan refers to the rope connected to the fuselage of the motorized capstan. It is mainly a rope that plays the role of balance, guidance and auxiliary braking during the operation of the motorized capstan. When the motorized capstan is working, especially when pulling heavier objects or tightening the rope, a large reaction force will be generated. If there is no tail rope to balance, the motorized capstan may move or even fall due to uneven force. For example, when the motorized capstan performs cable pulling operations on a hillside, due to the influence of gravity and the pulling direction, the motorized capstan itself will be subjected to a force opposite to the pulling direction. The tail rope can be pulled by manpower to keep the motorized capstan in a stable position. From a mechanical point of view, the traction force, self-weight, ground friction and tail rope tension of the motorized capstan constitute a force system. When the motorized capstan is working, the traction force tries to change the equilibrium state of the motorized capstan, while the tail rope tension and the traction force offset each other to a certain extent. Combined with the ground friction and its own gravity, the motorized capstan is in a balanced state to ensure normal operation. In addition, although the motorized capstan has a certain guide device to control the traction direction of the rope, in actual operation, especially in complex construction sites or encountering some special situations (such as encountering obstacles during traction and needing to change the traction direction), the guide device of the motorized capstan itself may not be able to accurately control the traction direction. At this time, the construction personnel can fine-tune the traction direction of the motorized capstan by operating the tail rope (it can also be understood as pulling the tail rope). For example, in the construction of power lines, when the cable is to be bypassed by insulators and other components on some poles and towers, it is necessary to accurately control the traction direction of the cable. The tail rope plays a key role in assisting the adjustment in this process. However, when the motorized capstan is working, the tail rope of the motorized capstan will be moved by the traction action of the motorized capstan. If the construction workers are in the rope loop of the tail rope of the motorized capstan, once the tail rope is subjected to a large tension or a sudden displacement, it is easy to entangle the construction workers. For example, when the motorized capstan is pulling the cable quickly or tightening the line, the tension of the tail rope will increase instantly. At this time, the rope loop will tighten rapidly, and the construction workers will have no time to dodge. Their bodies will be tightly entangled by the rope, thus falling into danger. In addition, due to the large traction force of the motorized capstan, the tail rope may move with the construction workers in the rope loop. In this case, the construction workers may be dragged to the ground and then dragged on the ground, causing skin abrasions, soft tissue contusions and other injuries, and may even be dragged to a dangerous area.In order to avoid such injuries, the construction worker operating the tail rope of the motorized capstan needs to be outside the rope loop of the tail rope of the motorized capstan.

[0085] In step S1, before the drone monitors the safety of the motorized capstan, the drone needs to fly above the motorized capstan to capture a bird's-eye view of the motorized capstan and construction personnel. The area above the motorized capstan is considered to be the first target wire-laying construction area, and a camera is provided at the bottom of the drone to capture a bird's-eye view of the motorized capstan, the tail rope of the motorized capstan, and construction personnel.

[0086] In step S3, the target detection model may specifically be YOLO (You Only Look Once), and the output of YOLO mainly includes a target box, the coordinates of the target box, and the category of the detected object; wherein the target box is a rectangular box, and the input image of YOLO is a rectangular image.

[0087] In step S5, if the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop intersect and the area of ​​the intersection is greater than the preset area threshold, then it can be determined that the construction worker operating the tail rope of the motorized capstan is located inside the rope loop of the tail rope of the motorized capstan, otherwise it can be determined that the construction worker operating the tail rope of the motorized capstan is located outside the rope loop of the tail rope of the motorized capstan. The area threshold is determined after analyzing and testing a large number of actual situations. It should be noted that the intersection area refers to the area of ​​intersection of the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop, that is, the common area.

[0088] In step S7, the step of calculating the area of ​​the rope loop includes: obtaining the length and width of the target frame of the rope loop, and calculating the area of ​​the rope loop using the length and width of the target frame and an ellipse calculation formula. It should be noted that in addition to this method for calculating the area of ​​the rope loop, other methods may also be used, which are not limited in the present embodiment of the application.

[0089] In step S9, the step of identifying the operating state of the motorized capstan includes: identifying the motorized capstan based on the target model and outputting the position information of the motorized capstan, and identifying the operating state of the motorized capstan according to the position information of the motorized capstan and through the infrared thermal imaging camera set on the drone; wherein the operating state of the motorized capstan includes an operating state and a non-operating state. Or the operating state of the motorized capstan is determined by analyzing the motion state of the motorized capstan components.

[0090] In step S10, specifically, the safety risk level is designed in such a way that the risk of the operating state is higher than the risk of the non-operating state and the higher the ratio of the area of ​​the intersection area to the rope loop, the higher the risk, and it is stored accordingly so that the safety risk level can be determined later according to the operating state of the motorized winch and the ratio of the area of ​​the intersection area to the rope loop.

[0091] In steps S11 and S12, in one embodiment, a warning strategy is determined according to the security risk level. For example, when the security risk level is level one (the higher the level, the greater the risk), the warning strategy may be to play a level one risk voice through a voice playback module on the drone, and the level one risk voice includes the current security risk level. When the security risk level is level two, a level two risk voice is played through a voice playback module on the drone, and the level one risk voice includes the current security risk level. In addition, the severity, urgency, etc. of the sound output by the voice module may be different for different security risk levels. The louder and more urgent the sound, the higher the security risk. At this time, the commander of the motorized winch can operate the motorized winch to reduce the risk, for example, the motorized winch can be turned off or slowed down.

[0092] The image obtained from the top view angle of the present application can more intuitively show the relative position relationship between the construction workers and the rope loop. Since the rope loop of the tail rope is usually on the ground, the top view image can more easily distinguish whether the construction workers are inside or outside the rope loop. This position relationship is clear at a glance from the top view angle of the plane. Compared with observing from the side or other angles, it reduces the judgment error caused by occlusion, etc., and improves the accuracy of detection. In addition, by comparing whether the target frame of the construction worker and the target frame of the rope loop have an intersection, the position relationship between the construction worker and the rope loop in space can be intuitively understood. This judgment method based on the intersection of the target frame simplifies the complex three-dimensional spatial position relationship into a geometric relationship judgment on a two-dimensional plane, reduces the difficulty and complexity of the judgment, and improves the speed of detection. When the target frame has an intersection, it means that the construction worker and the rope loop have overlapping parts on the image plane, and it is possible that the construction worker is located in the rope loop. On the basis of judging that there is an intersection, the area of ​​the intersection area is further calculated and compared with the preset area threshold, so that the judgment is more accurate. If the intersection area is small, it cannot be determined that the construction worker is inside the rope loop, for example, if the shoulder partially overlaps the rope loop; only when the intersection area is larger than the preset area threshold can it be more reliably determined that the construction worker is inside the rope loop, avoiding misjudgment caused by slight overlap of the target frame. In addition, by calculating the ratio of the intersection area to the rope loop area, it is possible to more accurately determine the proportion of the construction worker in the rope loop. Combined with the operating status of the motorized capstan, it is possible to more accurately determine the safety risk level of the construction worker. Different warning strategies are used according to different safety risk levels, which is more targeted and greatly protects the life safety and physical health of the construction workers.

[0093] In one embodiment, the position information of the target frame of the construction worker operating the tail rope of the motorized capstan includes the lower left corner coordinates (x1, y1) of the target frame of the construction worker operating the tail rope of the motorized capstan and the upper right corner coordinates (x2, y2) of the target frame of the construction worker operating the tail rope of the motorized capstan, and the position information of the target frame of the rope loop includes the lower left corner coordinates (x3, y3) of the target frame of the rope loop and the upper right corner coordinates (x4, y4) of the target frame of the rope loop;

[0094] The step of judging whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop intersects according to the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop comprises:

[0095] Determine whether max(x1,x3)<=min(x2,x4) and max(y1,y3)<=min(y2,y4) are satisfied; where min means taking the minimum value and max means taking the maximum value;

[0096] If so, it is determined that the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have an intersection; otherwise, it is determined that the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have no intersection.

[0097] In the embodiment of the present application, the method of judging whether the target frame of the construction worker operating the tail rope of the motorized winch and the target frame of the rope loop have an intersection is simple and can ensure a high recognition speed by judging whether max(x1,x3)<=min(x2,x4) and max(y1,y3)<=min(y2,y4).

[0098] In one embodiment, the step of calculating the area of ​​the intersection region includes:

[0099] The area of ​​the intersection region is calculated according to the following formula;

[0100] S=[min(x2,x4)-max(x1,x3)]×[min(y2,y4)-max(y1,y3)];

[0101] Where S is the area of ​​the intersection region.

[0102] In one embodiment, the step of identifying a rope loop of a tail rope of a motorized capstan based on the first image and the target detection model and outputting position information of a target frame of the rope loop comprises:

[0103] Identify the tail rope of the motorized capstan and the construction personnel through the target detection model, and output the position information of the tail rope of the motorized capstan and the position information of the construction personnel;

[0104] Calculate the shortest distance from the construction worker to the tail rope of the motorized capstan according to the position information of the tail rope of the motorized capstan and the position information of the construction worker;

[0105] Determining whether the shortest distance is less than a preset first threshold;

[0106] If it is less than, it is determined that the identified construction worker is a construction worker operating the tail rope of the motorized capstan. Otherwise, it is determined that the identified construction worker is not a construction worker operating the tail rope of the motorized capstan.

[0107] In the embodiment of the present application, since the images taken by the drone may include many construction workers, if the construction worker is far away from the tail rope, it is considered that the construction worker is not the construction worker operating the tail rope of the motorized capstan. If the construction worker is close to the tail rope and is right next to the tail rope, it can be considered that the construction worker is the construction worker operating the tail rope of the motorized capstan. This method is simple and can ensure a high recognition speed.

[0108] In one embodiment, the location information of the construction worker includes the center point coordinates of the construction worker target frame, the location information of the motorized capstan tail rope includes the location information of the motorized capstan tail rope target frame, and the step of calculating the shortest distance from the construction worker to the motorized capstan tail rope according to the location information of the motorized capstan tail rope and the location information of the construction worker includes:

[0109] Positioning the target frame of the motorized capstan tail rope according to the position information of the target frame of the motorized capstan tail rope;

[0110] Extracting the contour of the tail rope of the motorized capstan based on the target frame of the tail rope of the motorized capstan;

[0111] Calculate the distance between the center point coordinate of the construction worker target frame and each pixel point in the tail rope contour of the motorized capstan according to the center point coordinate of the construction worker target frame and the coordinate of each pixel point in the tail rope contour of the motorized capstan;

[0112] All the calculated distances are compared to obtain the shortest distance from the construction worker to the tail rope of the motorized capstan.

[0113] In an embodiment of the present application, by first accurately positioning the target frame according to the target frame position information of the motorized capstan tail rope, and then extracting the contour of the motorized capstan tail rope based on this, a foundation is laid for the subsequent accurate calculation of the distance from the construction personnel to the motorized capstan tail rope. Calculating the distance based on the coordinates of the center point of the construction personnel target frame and the coordinates of each pixel point of the motorized capstan tail rope contour can comprehensively consider the distance from the center point of the construction personnel target frame to each position on the tail rope contour, thereby obtaining a more accurate shortest distance value. The above-mentioned method for extracting the contour of the motorized capstan tail rope based on the target frame of the motorized capstan tail rope can adopt an edge detection algorithm, a contour tracking algorithm, or a morphological processing method.

[0114] In one embodiment, before identifying the tail rope of the motorized capstan and the construction personnel through the target detection model, the method further includes:

[0115] A large number of overhead images of the wiring construction area are taken by a drone; wherein the overhead image of the wiring construction area includes construction personnel, a motorized capstan, a tail rope of the motorized capstan, and a rope loop of the tail rope of the motorized capstan; wherein the overhead image is a rectangular image;

[0116] Classify and label each target object in the overhead image of the cable construction area to obtain a data set; wherein, classify and label each target object in the overhead image of the cable construction area includes classifying and labeling construction personnel, motorized capstan, motorized capstan tail rope, and rope loop of the motorized capstan tail rope; wherein, when labeling, the target frame is a rectangular frame, the target frame needs to frame the outline of each target object, and the four sides of the target frame are respectively parallel to the four sides of the overhead image;

[0117] The YOLO model is trained using the data set, and the training is stopped when the loss function converges to obtain the target detection model; wherein the target detection model is used to identify construction personnel, motorized capstans, motorized capstan tail ropes, and rope loops of motorized capstan tail ropes.

[0118] In an embodiment of the present application, the four sides of the target frame are respectively parallel to the four sides of the top view image, which means that when placed on the image, all its sides remain parallel to the corresponding sides of the image without rotation. When marking, manual marking is adopted. When marking, it is necessary to mark according to the requirement that the target frame frames the outline of the target object. Specifically, the marking is performed in a manner close to the outline, so that when the trained target detection model recognizes the object, the output target frame can more tightly frame the outline of the target object, so that it can more accurately judge whether the construction personnel operating the tail rope of the motorized winch are located in the rope loop of the tail rope of the motorized winch. The loss function includes classification loss, positioning loss and confidence loss. It should be understood that in the target detection task, the YOLO model needs to make category judgments on the objects in each detection frame. Classification loss is used to measure the error of the model in predicting the category of the object. Cross-Entropy Loss is usually used for calculation. Positioning loss is mainly used to measure the error between the target frame position predicted by the model and the true target frame position. Accurate target box positioning is a key part of target detection. The positioning loss function can help the model learn how to accurately determine the position and size of the object. Specifically, IoU loss can be used. Confidence loss: Confidence is used to indicate the model's confidence that the detected target box does contain the target object. It is usually calculated using cross entropy loss or binary cross entropy loss.

[0119] In one embodiment, after the step of determining whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop intersect according to the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop, the method further includes:

[0120] If the target frame of the construction worker operating the tail rope of the motorized capstan does not intersect with the target frame of the rope loop, then counting the number of identified construction workers operating the tail rope of the motorized capstan;

[0121] Determine whether the number of construction workers operating the tail rope of the motorized capstan is greater than 2;

[0122] If it is not greater than, a second voice is played through the voice playback module on the drone to warn the construction worker operating the tail rope of the motorized capstan that he has violated the rules and that it is necessary to ensure that there are more than two construction workers operating the tail rope of the motorized capstan.

[0123] In an embodiment of the present application, when it is identified that the number of construction workers is no more than 2, a second voice is promptly played through the voice playback module on the drone to issue a warning, reminding the construction workers that they have violated the regulations, prompting them to increase the number of operators to meet the requirements, thereby effectively preventing safety accidents such as loss of control of the winch due to insufficient number of construction workers, and ensuring the safety of the construction process.

[0124] In one embodiment, after the step of determining whether the number of construction workers operating the tail rope of the motorized capstan is greater than 2, the method further comprises:

[0125] If there are more than 2 persons, the motorized winch is identified based on the target detection model;

[0126] Calculating the distance between the motorized capstan and each construction worker operating the tail rope of the motorized capstan, and comparing them to obtain the distance between the construction worker closest to the motorized capstan and the motorized capstan;

[0127] Determining whether the distance between the construction worker closest to the motorized capstan and the motorized capstan is greater than a preset safety distance;

[0128] If it is less than, a third voice is played through the voice playback module on the drone to warn the construction worker operating the tail rope of the motorized capstan that he has violated the rules and to ensure that the distance between the construction worker closest to the motorized capstan and the motorized capstan is greater than the safety distance.

[0129] In the embodiment of the present application, it should be noted that, since the distance between the motorized capstan and each construction worker operating the tail rope of the motorized capstan is compared, the construction worker closest to the motorized capstan obtained above refers to the construction worker operating the tail rope of the motorized capstan closest to the motorized capstan. When the distance between the construction worker closest to the motorized capstan and the motorized capstan is less than the preset safety distance, the embodiment of the present application promptly plays a third voice through the voice playback module of the drone to issue a warning, prompting the construction worker to immediately adjust the position to ensure that the safety distance requirements are met, thereby effectively reducing the probability of safety accidents caused by distance violations and ensuring the personal safety of the construction workers.

[0130] In one embodiment, before the step of obtaining the overhead image of the first target cable construction area photographed by the drone, the method further includes:

[0131] Receive safety monitoring instructions.

[0132] Determining a safety monitoring mode according to the safety monitoring instruction;

[0133] When it is determined that the safety monitoring mode is the motorized capstan use safety monitoring mode, the step of obtaining the overhead view image of the first target wiring construction area photographed by the drone is executed.

[0134] In an embodiment of the present application, a user sends a safety monitoring instruction to a drone through a drone remote controller. After receiving the safety monitoring instruction, the drone determines a safety monitoring mode according to the safety monitoring instruction. Specifically, the safety monitoring instruction carries an identifier, and the drone determines a specific safety monitoring mode according to the relationship between the identifier and the safety monitoring mode; wherein one identifier corresponds to one safety monitoring mode, wherein one safety monitoring mode is a motorized capstan use safety monitoring mode, and other safety monitoring modes may be seat belt wearing monitoring modes, etc. Before the drone is required to monitor the safety of the motorized capstan use, it is necessary to control the drone to fly above the motorized capstan to capture a bird's-eye view of the motorized capstan and construction personnel. Among them, the area above the motorized capstan is considered to be the first target wireline construction area.

[0135] In one embodiment, before the step of obtaining the overhead image of the first target cable construction area photographed by the drone, the method further includes:

[0136] Obtain inspection routes for cable construction areas;

[0137] During the inspection process according to the inspection route, when the motorized capstan and the construction personnel are identified at the same time, the drone turns on the motorized capstan to use the safety monitoring mode, and executes the step of obtaining the overhead image of the first target wire construction area taken by the drone.

[0138] In the embodiment of the present application, automatic inspection is used to reduce user operations and is more intelligent.

[0139] In one embodiment, the method further comprises: warning the construction worker who operates the tail rope of the motorized capstan according to the warning strategy through the voice playback module of the drone; wherein the warning is used to warn the construction worker who operates the tail rope of the motorized capstan not to stand in the rope loop.

[0140] The human body posture of the construction worker operating the tail rope of the motorized capstan is estimated by a posture estimation algorithm, and whether the construction worker operating the tail rope of the motorized capstan is in a falling state is determined according to the human body posture estimation result;

[0141] If it is determined that the machine is in a falling state, the drone sends a command to shut down or slow down the motorized capstan to the commander of the motorized capstan through the voice playback module, or sends a command to shut down or slow down the motorized capstan to the communication device of the commander of the motorized capstan through the communication module.

[0142] In the embodiment of the present application, specifically, a posture estimation algorithm can be used to detect whether the central axis of the body is close to being parallel to the ground. If so, it is determined that the construction worker operating the tail rope of the motorized winch is in a fallen state.

[0143] In the embodiment of the present application, when it is detected that the construction worker operating the tail rope of the motorized capstan is located in the rope loop, a human posture estimation algorithm is used to detect whether the construction worker operating the tail rope of the motorized capstan is in a falling state. When the construction worker operating the tail rope of the motorized capstan is in a falling state, it indicates that the construction worker operating the tail rope of the motorized capstan is trapped and dragged by the rope loop. At this time, the drone sends a command to shut down or slow down the motorized capstan to the commander of the motorized capstan through the voice playback module, or sends a command to shut down or slow down the motorized capstan to the communication device of the commander of the motorized capstan through the communication module, which can effectively prevent the construction worker from suffering more serious injuries due to being dragged by the rope loop. The commander can also be identified by training the YOLO model.

[0144] The present application also provides a wireline construction safety monitoring system based on image processing, which is applied to a drone and includes:

[0145] An image acquisition module is used to acquire a bird's-eye view image of the first target cable construction area photographed by the drone; wherein the bird's-eye view image includes a motorized capstan, a tail rope of the motorized capstan, and construction personnel;

[0146] An equivalent module, configured to use the overhead image of the first target wiring construction area photographed by the drone as the first image;

[0147] A first recognition module is used to identify the construction worker who operates the tail rope of the motorized capstan based on the first image, the target detection model and the distance calculation method, and output the position information of the target frame of the construction worker who operates the tail rope of the motorized capstan; wherein the target frame output by the target detection model is a rectangular frame, the image input by the target detection model is a rectangular image, and the four sides of the target frame are respectively parallel to the four sides of the input image;

[0148] A second recognition module, which recognizes the rope loop of the tail rope of the motorized capstan based on the first image and the target detection model and outputs the position information of the target frame of the rope loop;

[0149] A first judgment module is used to judge whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have an intersection according to the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop;

[0150] A first calculation module is used to calculate the area of ​​the intersection region, if any;

[0151] A second calculation module, configured to calculate the area of ​​the rope loop if the area of ​​the intersection region is greater than a preset area threshold;

[0152] A third calculation module, configured to obtain a ratio of the area of ​​the intersection area to the area of ​​the rope loop by dividing the area of ​​the intersection area by the area of ​​the rope loop;

[0153] A third identification module, used for identifying the operating state of the motorized winch;

[0154] A safety risk level determination module, used to determine the safety risk level according to the operating state of the motorized capstan and the ratio of the intersection area to the area of ​​the rope loop;

[0155] A warning strategy determination module, used to determine a warning strategy according to the security risk level;

[0156] The warning module is used to warn the construction personnel who operate the tail rope of the motorized capstan according to the warning strategy through the voice playback module of the drone; wherein the warning is used to warn the construction personnel who operate the tail rope of the motorized capstan not to stand in the rope loop.

[0157] In one embodiment, the position information of the target frame of the construction worker operating the tail rope of the motorized capstan includes the lower left corner coordinates (x1, y1) of the target frame of the construction worker operating the tail rope of the motorized capstan and the upper right corner coordinates (x2, y2) of the target frame of the construction worker operating the tail rope of the motorized capstan, and the position information of the target frame of the rope loop includes the lower left corner coordinates (x3, y3) of the target frame of the rope loop and the upper right corner coordinates (x4, y4) of the target frame of the rope loop;

[0158] The step of judging whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop intersects with each other according to the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop comprises:

[0159] Determine whether max(x1,x3)<=min(x2,x4) and max(y1,y3)<=min(y2,y4) are satisfied;

[0160] If so, it is determined that the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have an intersection; otherwise, it is determined that the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have no intersection.

[0161] It should be noted that the wireline construction safety monitoring method based on image processing of the present invention is applicable to the wireline construction safety monitoring system based on image processing of the present invention, and therefore, the present invention will not be described in detail herein.

[0162] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided in this application and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0163] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0164] The above description is only a preferred embodiment of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A wireline construction safety monitoring method based on image processing, characterized in that: The method is applied to a drone, comprising: Acquire a bird's-eye view image of the first target cable construction area photographed by the drone; wherein the bird's-eye view image includes a motorized capstan, a tail rope of the motorized capstan, and construction personnel; Using a bird's-eye view image of the first target wiring construction area photographed by the drone as a first image; Based on the first image, the target detection model and the distance calculation method, the construction worker operating the tail rope of the motorized capstan is identified and the position information of the target frame of the construction worker operating the tail rope of the motorized capstan is output; wherein the target frame output by the target detection model is a rectangular frame, the image input by the target detection model is a rectangular image, and the four sides of the target frame are respectively parallel to the four sides of the input image; Identify the rope loop of the tail rope of the motorized capstan based on the first image and the target detection model and output the position information of the target frame of the rope loop; Determining whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop intersect according to the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop; If so, calculate the area of ​​the intersection; If the area of ​​the intersection region is greater than a preset area threshold, then calculating the area of ​​the rope loop; Dividing the area of ​​the intersection region by the area of ​​the rope loop to obtain a ratio of the area of ​​the intersection region to the area of ​​the rope loop; Identifying the operating status of the motorized capstan; Determining the safety risk level according to the operating state of the motorized capstan and the ratio of the intersection area to the area of ​​the rope loop; Determining a warning strategy based on the security risk level; The construction personnel operating the tail rope of the motorized capstan are warned according to the warning strategy through the voice playback module of the drone; wherein the warning is used to warn the construction personnel operating the tail rope of the motorized capstan not to stand in the rope loop.

2. The wireline construction safety monitoring method based on image processing according to claim 1 is characterized in that: The position information of the target frame of the construction worker operating the tail rope of the motorized capstan includes the lower left corner coordinates (x1, y1) of the target frame of the construction worker operating the tail rope of the motorized capstan and the upper right corner coordinates (x2, y2) of the target frame of the construction worker operating the tail rope of the motorized capstan, and the position information of the target frame of the rope loop includes the lower left corner coordinates (x3, y3) of the target frame of the rope loop and the upper right corner coordinates (x4, y4) of the target frame of the rope loop; The step of judging whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop intersects according to the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop comprises: Determine whether max(x1,x3)<=min(x2,x4) and max(y1,y3)<=min(y2,y4) are satisfied; where min means taking the minimum value and max means taking the maximum value; If so, it is determined that the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have an intersection; otherwise, it is determined that the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have no intersection.

3. The wireline construction safety monitoring method based on image processing according to claim 2 is characterized in that: The step of calculating the area of ​​the intersection region comprises: The area of ​​the intersection region is calculated according to the following formula; S=[min(x2,x4)-max(x1,x3)]×[min(y2,y4)-max(y1,y3)]; Where S is the area of ​​the intersection region.

4. The wireline construction safety monitoring method based on image processing according to claim 1 is characterized in that: The step of identifying the rope loop of the tail rope of the motorized capstan based on the first image and the target detection model and outputting the position information of the target frame of the rope loop comprises: Identify the tail rope of the motorized capstan and the construction personnel through the target detection model, and output the position information of the tail rope of the motorized capstan and the position information of the construction personnel; Calculate the shortest distance from the construction worker to the tail rope of the motorized capstan according to the position information of the tail rope of the motorized capstan and the position information of the construction worker; Determining whether the shortest distance is less than a preset first threshold; If it is less than, it is determined that the identified construction worker is a construction worker operating the tail rope of the motorized capstan. Otherwise, it is determined that the identified construction worker is not a construction worker operating the tail rope of the motorized capstan.

5. The wireline construction safety monitoring method based on image processing according to claim 4 is characterized in that: The position information of the construction worker includes the center point coordinates of the construction worker target frame, the position information of the motorized capstan tail rope includes the position information of the motorized capstan tail rope target frame, and the step of calculating the shortest distance from the construction worker to the motorized capstan tail rope according to the position information of the motorized capstan tail rope and the position information of the construction worker includes: Positioning the target frame of the motorized capstan tail rope according to the position information of the target frame of the motorized capstan tail rope; Extracting the contour of the tail rope of the motorized capstan based on the target frame of the tail rope of the motorized capstan; Calculate the distance between the center point coordinate of the construction worker target frame and each pixel point in the tail rope contour of the motorized capstan according to the center point coordinate of the construction worker target frame and the coordinate of each pixel point in the tail rope contour of the motorized capstan; All the calculated distances are compared to obtain the shortest distance from the construction worker to the tail rope of the motorized capstan.

6. The wireline construction safety monitoring method based on image processing according to claim 4 is characterized in that: Before the tail rope of the motorized capstan and the construction personnel are identified by the target detection model, the method further includes: A large number of overhead images of the wiring construction area are taken by a drone; wherein the overhead image of the wiring construction area includes construction personnel, a motorized capstan, a tail rope of the motorized capstan, and a rope loop of the tail rope of the motorized capstan; wherein the overhead image is a rectangular image; Classify and label each target object in the overhead image of the cable construction area to obtain a data set; wherein, classify and label each target object in the overhead image of the cable construction area includes classifying and labeling construction personnel, motorized capstan, motorized capstan tail rope, and rope loop of the motorized capstan tail rope; wherein, when labeling, the target frame is a rectangular frame, the target frame needs to frame the outline of each target object, and the four sides of the target frame are respectively parallel to the four sides of the overhead image; The YOLO model is trained using the data set, and the training is stopped when the loss function converges to obtain the target detection model; wherein the target detection model is used to identify construction personnel, motorized capstans, motorized capstan tail ropes, and rope loops of motorized capstan tail ropes.

7. The wireline construction safety monitoring method based on image processing according to claim 1 is characterized in that: After the step of judging whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop intersect according to the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop, the method further comprises: If the target frame of the construction worker operating the tail rope of the motorized capstan has no intersection with the target frame of the rope loop, then counting the number of construction workers identified to operate the tail rope of the motorized capstan; Determine whether the number of construction workers operating the tail rope of the motorized capstan is greater than 2; If it is not greater than, a second voice is played through the voice playback module on the drone to warn the construction worker operating the tail rope of the motorized capstan that he has violated the rules and that it is necessary to ensure that there are more than two construction workers operating the tail rope of the motorized capstan.

8. The method for monitoring the safety of wireline construction based on image processing according to claim 7, characterized in that: After the step of determining whether the number of construction workers operating the tail rope of the motorized capstan is greater than 2, the method further comprises: If there are more than 2 persons, the motorized winch is identified based on the target detection model; Calculating the distance between the motorized capstan and each construction worker operating the tail rope of the motorized capstan, and comparing them to obtain the distance between the construction worker closest to the motorized capstan and the motorized capstan; Determining whether the distance between the construction worker closest to the motorized capstan and the motorized capstan is greater than a preset safety distance; If it is less than, a third voice is played through the voice playback module on the drone to warn the construction worker operating the tail rope of the motorized capstan that he has violated the rules and to ensure that the distance between the construction worker closest to the motorized capstan and the motorized capstan is greater than the safety distance.

9. A wireline construction safety monitoring system based on image processing, characterized in that: The system is applied to a drone, and includes: An image acquisition module is used to acquire a bird's-eye view image of the first target cable construction area photographed by the drone; wherein the bird's-eye view image includes a motorized capstan, a tail rope of the motorized capstan, and construction personnel; An equivalent module, configured to use the overhead image of the first target wiring construction area photographed by the drone as the first image; A first recognition module is used to identify the construction worker who operates the tail rope of the motorized capstan based on the first image, the target detection model and the distance calculation method, and output the position information of the target frame of the construction worker who operates the tail rope of the motorized capstan; wherein the target frame output by the target detection model is a rectangular frame, the image input by the target detection model is a rectangular image, and the four sides of the target frame are respectively parallel to the four sides of the input image; A second recognition module, which recognizes the rope loop of the tail rope of the motorized capstan based on the first image and the target detection model and outputs the position information of the target frame of the rope loop; A first judgment module is used to judge whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have an intersection according to the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop; A first calculation module is used to calculate the area of ​​the intersection region, if any; A second calculation module, configured to calculate the area of ​​the rope loop if the area of ​​the intersection region is greater than a preset area threshold; A third calculation module, configured to obtain a ratio of the area of ​​the intersection area to the area of ​​the rope loop by dividing the area of ​​the intersection area by the area of ​​the rope loop; A third identification module, used for identifying the operating state of the motorized winch; A safety risk level determination module, used to determine the safety risk level according to the operating state of the motorized capstan and the ratio of the intersection area to the area of ​​the rope loop; A warning strategy determination module, used to determine a warning strategy according to the security risk level; The warning module is used to warn the construction personnel who operate the tail rope of the motorized capstan according to the warning strategy through the voice playback module of the drone; wherein the warning is used to warn the construction personnel who operate the tail rope of the motorized capstan not to stand in the rope loop.

10. The wireline construction safety monitoring system based on image processing according to claim 9, characterized in that: The position information of the target frame of the construction worker operating the tail rope of the motorized capstan includes the lower left corner coordinates (x1, y1) of the target frame of the construction worker operating the tail rope of the motorized capstan and the upper right corner coordinates (x2, y2) of the target frame of the construction worker operating the tail rope of the motorized capstan, and the position information of the target frame of the rope loop includes the lower left corner coordinates (x3, y3) of the target frame of the rope loop and the upper right corner coordinates (x4, y4) of the target frame of the rope loop; The step of judging whether the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop intersects with each other according to the position information of the target frame of the construction worker operating the tail rope of the motorized capstan and the position information of the target frame of the rope loop comprises: Determine whether max(x1,x3)<=min(x2,x4) and max(y1,y3)<=min(y2,y4) are satisfied; If so, it is determined that the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have an intersection; otherwise, it is determined that the target frame of the construction worker operating the tail rope of the motorized capstan and the target frame of the rope loop have no intersection.

Images