A method and system for detecting the suspension status of safety rope hooks used in high-altitude operations

Abstract

This invention discloses a method and system for detecting the suspension status of safety rope hooks in high-altitude operations, relating to the field of image recognition technology. The method includes: determining a candidate set of stable regions based on the distribution density and time interval changes of the image center point within each grid area, and selecting a target suspension region based on the candidate set; acquiring the current hook image after obtaining the target suspension region, extracting the hook edge contour features based on the hook image, and calculating the hook's current spatial attitude parameters; comparing the current spatial attitude parameters with a preset standard hook attitude range to determine whether the current spatial attitude parameters belong to the standard hook attitude range; and calculating an attitude adjustment vector when the current spatial attitude parameters do not belong to the standard hook attitude range. This invention achieves real-time and accurate detection and active attitude correction of the safety rope hook suspension status, effectively ensuring the safety of high-altitude workers.

Description

Technical Field

[0001] This invention relates to the field of image recognition technology, specifically to a method and system for detecting the suspension status of safety rope hooks used in high-altitude operations. Background Technology

[0002] With the continuous expansion of high-altitude operations and increasingly stringent safety standards, safety rope hooks, as crucial connection devices for ensuring safety in high-altitude operations, directly impact the personal safety of workers through the accuracy and stability of their suspension status. However, in practical engineering applications, the detection of the suspension status of safety rope hooks often relies on the subjective judgment of workers based on their experience. This method not only fails to achieve accurate quantitative assessment but is also easily affected by human factors such as insufficient experience, fatigue, and obstructed vision, leading to the overlooking or misjudgment of abnormal suspension states. Furthermore, due to the complex and variable working environment at heights, external factors such as wind interference, vibration of the work platform, and the hook's own swaying can cause imperceptible deviations and abnormalities in the hook's position and posture. Failure to detect and correct these deviations in a timely manner can easily lead to serious safety accidents.

[0003] While existing automated video monitoring solutions offer certain advantages in automation, they generally suffer from the following shortcomings: First, they lack real-time, detailed analysis of the dynamic motion trajectory of the hook, making it difficult to accurately quantify the hook's true spatial position and motion stability. Second, hook posture detection is mostly limited to static image analysis, failing to fully utilize the spatial orientation information of the hook's edge contour for in-depth analysis, resulting in low posture recognition accuracy. Third, existing technologies have failed to construct effective quantitative evaluation standards and active correction mechanisms for hook posture anomalies, meaning that even if posture anomalies are detected, real-time active correction is not possible, requiring manual intervention, which is slow and poses safety hazards. Summary of the Invention

[0004] The purpose of this invention is to provide a safety rope hook suspension status detection system and method for high-altitude operations, so as to solve the problems in the background art mentioned above.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] In a first aspect, the present invention provides a method for detecting the suspension status of a safety rope hook used in high-altitude operations, characterized in that it includes:

[0007] The sequence of center points of the acquired images is mapped to the coordinate space of the target surface. Based on the distribution density and time interval of the center points of the images in each grid area, a candidate set of stable regions is determined, and the target hanging region is selected according to the candidate set of stable regions.

[0008] After acquiring the target suspension area, the current hook image is captured. Based on the hook image, the hook edge contour features are extracted, and the current spatial attitude parameters of the hook are calculated. The spatial attitude parameters include the tilt angle and rotation direction angle of the hook's main edge line segment.

[0009] The current spatial attitude parameters are compared with the preset hook standard attitude range to determine whether the current spatial attitude parameters belong to the hook standard attitude range.

[0010] When the current spatial attitude parameters do not belong to the standard attitude range of the hook, calculate the attitude adjustment vector, and guide the hook correction actuator to perform the correction operation on the hook attitude based on the attitude adjustment vector.

[0011] Secondly, the present invention provides a safety rope hook suspension status detection system for high-altitude operations, implemented based on the aforementioned safety rope hook suspension status detection method for high-altitude operations, characterized in that it includes:

[0012] The acquisition module is used to map the sequence of center points of the acquired images to the coordinate space of the target surface, and to determine the candidate set of stable regions based on the distribution density and time interval changes of the image center points in each grid area, and to filter out the target hanging region based on the candidate set of stable regions.

[0013] The extraction module is used to acquire the current hook image after obtaining the target suspension area, extract the hook edge contour features based on the hook image, and calculate the current spatial attitude parameters of the hook. The spatial attitude parameters include the tilt angle and rotation direction angle of the hook's main edge line segment.

[0014] The comparison module is used to compare the current spatial attitude parameters with the preset hook standard attitude range to determine whether the current spatial attitude parameters belong to the hook standard attitude range.

[0015] The control module is used to calculate the attitude adjustment vector when the current spatial attitude parameters do not belong to the standard attitude range of the hook, and guide the hook correction actuator to perform the correction operation on the hook attitude based on the attitude adjustment vector.

[0016] Thirdly, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the above-described method for detecting the suspension status of a safety rope hook for high-altitude operations.

[0017] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed, implements the above-described method for detecting the suspension status of a safety rope hook for high-altitude operations.

[0018] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0019] This invention achieves accurate quantification and stability determination of the hook's suspension area and spatial posture by real-time acquisition of video image sequences of the safety rope hook and by using spatial mapping and gridding analysis methods. This effectively solves the problems of insufficient accuracy in hook position and posture detection and susceptibility to human and environmental interference in the prior art, and enables early identification and accurate positioning of abnormal hook states.

[0020] Furthermore, by extracting the edge contour features of the hook and calculating the spatial attitude parameters of the hook edge line segment direction, an objective and accurate standard attitude range for the hook is constructed. This overcomes the limitations of existing methods that rely on subjective experience judgment, enabling the identification of abnormal hook attitude states to be more precise and standardized, and improving the accuracy and reliability of abnormal state identification.

[0021] Furthermore, by calculating the hook attitude adjustment vector and using real-time quantified attitude deviation as a basis, the automatic correction actuator is guided to actively correct abnormal attitudes, which significantly improves the timeliness and effectiveness of hook suspension state correction, reduces safety risks caused by hook abnormalities during high-altitude operations, and protects the personal safety of workers, thus having significant practical application value. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0023] Figure 1 This is a flowchart of a method for detecting the suspension status of a safety rope hook used in high-altitude operations according to the present invention;

[0024] Figure 2 This is a framework diagram of a safety rope hook suspension status detection system for high-altitude operations according to the present invention. Detailed Implementation

[0025] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided to make the description of this application more complete and comprehensive, and to fully convey the concept of the exemplary embodiments to those skilled in the art. The drawings are merely illustrative illustrations of this application and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted.

[0026] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more exemplary embodiments. Numerous specific details are provided in the following description to give a full understanding of the exemplary embodiments disclosed in this application. However, those skilled in the art will recognize that the technical solutions disclosed in this application can be practiced with one or more specific details omitted, or other methods, components, steps, etc., can be employed. In other instances, well-known structures, methods, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of the disclosure of this application.

[0027] Example 1

[0028] like Figure 1 As shown, this embodiment discloses a method for detecting the suspension status of a safety rope hook used in high-altitude operations, including:

[0029] S101: Map the sequence of center points of the acquired images to the coordinate space of the target surface, and determine the candidate set of stable regions based on the distribution density and time interval of the image center points in each grid area, and select the target hanging region according to the candidate set of stable regions.

[0030] It should be noted that the video image sequence, including multiple consecutive video images, is acquired by a fixed high-definition camera installed on the safety belt or safety helmet. The camera has a frame rate of no less than 30 frames per second, which can cover the entire field of view of the hook movement and continuously acquire the image frame sequence of the entire hook movement process.

[0031] In implementation, determining the candidate set of stability regions includes:

[0032] The target surface coordinate space is divided into multiple grid regions of fixed size, and each image center point in the image center point sequence is projected onto the target surface coordinate space and assigned to its corresponding grid region.

[0033] In practice, the acquired video image sequence is first processed frame by frame. The center point of the hook area is extracted from each frame to form a sequence of image center points, which is used to represent the temporal trajectory of the hook in the image time axis.

[0034] It should be noted that the image center point is the geometric center point of the smallest bounding rectangle corresponding to the hook detection box (or segmentation mask), and the calculation formula is:

[0035]

[0036] Where, x min ,x max ,y min ,ymax These are the boundary coordinates of the detection box in the image coordinate system, in pixels;

[0037] Next, the center point of the image is projected from the image pixel coordinate system to the world coordinate system (i.e., the target surface coordinate space) where the working surface is located, in order to obtain the position of the hook in the real physical space.

[0038] It should be noted that the camera calibration process uses Zhang Zhengyou's method to jointly obtain the camera's intrinsic and extrinsic parameters. Through the intrinsic parameter matrix K, rotation matrix R, and displacement vector T, the transformation relationship from image pixel coordinates to physical space coordinates can be established.

[0039] The center point of each frame's image (x) c ,y c The three-dimensional spatial point (X) is obtained through the following projection relationship transformation. c ,Y c Z c Then project it onto the target surface coordinate plane Z = Z0 to obtain the plane coordinates (X). t ,Y t ):

[0040]

[0041] Where s is the scaling factor, Z c Given a known depth (set to a fixed height between the working plane and the camera);

[0042] After completing the coordinate projection, a two-dimensional planar mesh system is established in the coordinate space of the target surface. The mesh size is set to a fixed value, for example, each mesh is 50mm × 50mm.

[0043] For example, if the entire working surface area is 1000mm × 1000mm, it can be divided into 20 × 20 numbered grid areas, numbered as G. i,j , where i and j are the horizontal and vertical grid indices, respectively, which increase sequentially starting from the top left corner;

[0044] Specifically, for each center point (X) of the projected hook image t ,Y t The corresponding grid number is determined in the following way:

[0045] Set the origin of the coordinate system at the top left corner of the work surface;

[0046] Calculate the horizontal number of the grid in which it is located:

[0047] Calculate the vertical numbering:

[0048] Among them, Lg The side length of a single grid cell (e.g., 50 mm);

[0049] Finally, each image center point is assigned to its corresponding grid number G. i,j This forms a time-space sequence:

[0050] For each grid region, count the number of image center points assigned to that grid region, and calculate the distribution density value obtained by dividing the number of image center points in that grid region by the image acquisition time;

[0051] Specifically, for each grid region G i,j Traverse the image center point-grid number correspondence sequence Count the number of image frame numbers that fall within this grid region, and denote it as N. i,j ;

[0052] It should be noted that this statistical operation is based on frame number and does not depend on specific inter-frame time, and can directly reflect the total number of times the hook appears in this area;

[0053] Then, obtain the acquisition duration T of the current video image, in seconds. If the total number of video frames is F and the camera frame rate is fps, then the calculation formula is:

[0054]

[0055] The formula for calculating the distribution density value is as follows:

[0056]

[0057] Among them, D i,j Represents grid G i,j The distribution density value;

[0058] It should be noted that the density value reflects the frequency of hooks appearing in the region per unit time. The larger the value, the stronger the spatial behavior stability of the region, and it can be used as one of the reference indicators for candidate target regions.

[0059] The final output is the set of distribution density values ​​corresponding to each grid number:

[0060] For each grid region, extract the image frame number sequence of the image center point distributed in the region, calculate the set of differences between adjacent frame numbers in the sequence, and perform a range-to-mean ratio operation on the set of differences to obtain the time interval variation coefficient of the region.

[0061] Specifically, for each grid region G i,j From the image center point - grid number corresponding sequence Filter out all frame numbers f t Its corresponding grid number satisfies This is denoted as the frame number sequence:

[0062] F i,j ={f1,f2,...,f n};

[0063] Among them, f k <f k+1 Sort by time;

[0064] Subsequently, the interval sequence between adjacent frame numbers is calculated: ΔF i,j ={f2-f1,f3-f2,...,f n -f n-1 This interval set reflects the time interval between each occurrence of the hook within the region;

[0065] It should be noted that the video frame rate (fps) can be used to convert the inter-frame interval into a time interval in seconds. For example, if the frame interval is 5 frames, the corresponding time interval is 5 / fps.

[0066] Next, the range and mean of the inter-frame interval sequence are calculated, and the time interval variation coefficient V is defined. i,j :

[0067]

[0068] It should be noted that if the time interval between each occurrence of the hook in this area is relatively uniform (i.e., the interval fluctuation is small), then the range is close to the mean, and the coefficient of variation V is... i,j The coefficient of variation approaches 0; conversely, if the hook only occasionally crosses a certain area and the time interval fluctuates greatly, the coefficient of variation will be greater.

[0069] The final output is a pair of grid numbers and their corresponding time interval variation coefficients:

[0070] Grid regions that simultaneously satisfy the conditions of having a distribution density value greater than or equal to the distribution density determination threshold and a time interval change coefficient less than or equal to the time interval change determination threshold will be included in the candidate set of stable regions.

[0071] Specifically, the distribution density set With the set of time interval variation coefficients Perform joint processing;

[0072] In implementation, a distribution density determination threshold D is set. min The threshold V for determining the change coefficient of the time interval maxThese two thresholds are system parameters, obtained through actual testing or experience.

[0073] Specifically, for each grid G ij If the following two conditions are met:

[0074]

[0075] Then the grid G i,j The candidate set of stable regions is included, and the final output is the candidate set of stable regions:

[0076] Specifically, the step of filtering the target dangling region based on the candidate set of stability regions includes:

[0077] In the time series of the image center points, the corresponding frame number of each image center point in the time sequence is extracted, and the grid region number to which each image center point belongs is recorded in sequence.

[0078] Specifically, based on the previously completed image center point spatial projection and grid assignment processing, the following sequence is constructed:

[0079]

[0080] Where: f t g represents the image frame number of frame t. t This indicates the grid number corresponding to the center point of the image frame;

[0081] In implementation, this time-grid number sequence will be used as input to the sliding window continuity judgment algorithm to identify whether the hook has remained unchanged in a certain region of the stability region candidate set for a period of time. The final output is a complete hook image center point frame number-grid number pair sequence:

[0082]

[0083] The consistency of the grid region numbers to which multiple consecutive image center points belong is judged. If the grid region numbers to which the consecutive image center points belong are consistent with the grid region numbers in the same set of candidate stable regions, then it is determined that the image center points stay continuously in that grid region.

[0084] It should be noted that "continuous dwell" means that the hook remains within the same grid number across multiple consecutive image frames, and that grid number belongs to the aforementioned candidate set S of stable regions. stable ;

[0085] Specifically, based on the aforementioned extracted image center point frame number – grid number sequence

[0086] In implementation, a fixed-length sliding window W is used to traverse the sequence and determine whether the grid numbers corresponding to each consecutive frame number are consistent. The specific steps are as follows:

[0087] Set the sliding window length W in frames, for example, set it to 30 frames (which corresponds to 1 second of video duration);

[0088] For any consecutive frame interval (f) t ,...,f t+W-1 Extract the corresponding grid number subsequence:

[0089]

[0090] Determine whether the following two conditions are met simultaneously:

[0091] All grid numbers g k They are all the same, that is, there exists a certain G. i,j , making

[0092] Grid G i,j Candidate set S belonging to the stable region stable ;

[0093] If both of the above conditions are met, then the hook is considered to be in region G. i,j A continuous dwelling behavior occurred, and the system entered the target area candidate state;

[0094] The output is: a set of region IDs that satisfy the condition of continuous stay.

[0095]

[0096] Each of them And it satisfies spatial consistency within a certain period of time;

[0097] When the length of the frame number sequence in which the center point of the image stays continuously in a grid region in a certain set of candidate stable regions meets the preset dwell length determination condition, the grid region is marked as the target hanging region.

[0098] Specifically, for the aforementioned set S of grid regions exhibiting "continuous dwelling behavior" stay Further, the duration is determined, specifically whether the number of frames the hook continuously stays in that area within each continuous pause meets the pause frame threshold N. min ;

[0099] It should be noted that: the threshold for the number of frames to remain is N. minThese are engineering parameters set based on the image frame rate to ensure that the hook has sufficient time to remain still within the target area to reflect the true stable hanging state. For example, at a frame rate of 30 frames per second, if the hook is required to remain still for 1 second, then N should be set. mim =30;

[0100] Specifically, traversing the time-grid number mapping sequence For each candidate region G k ∈S stay Find the number of consecutive frames N that appear. k If N satisfies: k ≥N min Then the grid region G k Marked as the final target suspension area G target ;

[0101] S102: After acquiring the target suspension area, acquire the current hook image, extract the hook edge contour features based on the hook image, and calculate the current spatial attitude parameters of the hook. The spatial attitude parameters include the tilt angle and rotation direction angle of the hook's main edge line segment.

[0102] In implementation, the extraction of hook edge contour features includes:

[0103] Extract the set of pixels whose brightness changes are greater than the change in the global average brightness of the image from the hook image, and construct an image brightness change map;

[0104] Specifically, for the input image frame I, the image is first converted into a grayscale image I. gray Then, the brightness variation regions of the grayscale image are extracted. The main processing flow is as follows:

[0105] For image I gray The gradient intensity G(x,y) of each pixel in the image is calculated by applying a first-order difference filter (such as the Sobel operator) or the Laplacian operator.

[0106] The gradient intensity of all pixels in the image is used to form a gradient map G;

[0107] Set gradient strength threshold T g Filter out those that satisfy G(x,y)≥T g The pixels are used to construct an initial brightness variation region mask M. edge

[0108] Record the positions of all non-zero pixels in the mask as the set of pixels with brightness changes;

[0109] Finally, the output image brightness variation map is a binary image M. edgeThe location with a pixel value of 1 represents an area with significant grayscale changes, while the location with a pixel value of 0 represents the background or non-edge area.

[0110] In the image brightness variation map, the set of continuous pixel paths is identified according to the connection relationship of pixel positions, and in each pixel path, the main direction angle sequence from the starting point to the ending point of the path is extracted;

[0111] It should be noted that the main direction angle sequence refers to the set of local direction angles calculated along the pixel path from the starting point to the ending point in the order of pixel connection. It is used to reflect the overall direction change trend of the path. This sequence is subsequently used to evaluate the path direction consistency and attitude angle calculation.

[0112] Specifically, the identification of the set of continuous pixel paths includes:

[0113] Traversing the image brightness variation graph M edge All edge pixels with a pixel value of 1;

[0114] For each unvisited pixel, an eight-neighbor pixel connection method is used to search for connected pixels in adjacent directions to construct a pixel connection path;

[0115] It should be noted that: eight-neighbor connectivity means that each pixel is considered to be adjacent to pixels in eight directions around it, which is convenient for extracting curved contours or irregular edges;

[0116] During the construction process, the image coordinates (x, y) of each pixel in the path are recorded. i ,y i ), forming a path sequence P k ={(x1,y1),...,(x m ,y n )};

[0117] Mark all pixels in the path as "visited" and store the path P. k ;

[0118] Repeat the above operation until all edge pixels are assigned to a path or have been visited, thus obtaining a set of continuous pixel paths;

[0119] Specifically, the extraction of the principal direction angle sequence from the starting point to the ending point of the path includes:

[0120] For path P k Each pair of adjacent pixels (x) i ,y i ),(x i+1 ,y i+1 ), calculate the direction angle θ i :

[0121] θi =arctan 2(y i+1 -y i ,x i+1 -x i );

[0122] Angle values ​​are uniformly converted to the interval [-π,π] or [0°,360°].

[0123] Obtain the sequence of principal direction angles for the current path: Θ k ={θ1,θ2,...,θ n-1};

[0124] The final output is a set of direction angle sequences for each path: {(P1,Θ1),(P2,Θ2),...,(P m ,Θ m )};

[0125] Calculate the standard deviation of the main direction angle sequence for each pixel path. If the standard deviation is less than the direction fluctuation threshold, then mark the path as a hook edge segment.

[0126] Specifically, for each path P k Perform the following directional volatility calculation procedure:

[0127] Calculate the mean of the angle sequence:

[0128]

[0129] Calculate the squared deviation of the angle in each direction from the mean:

[0130]

[0131] Calculate the standard deviation of the direction angle (i.e., the directional fluctuation):

[0132]

[0133] Furthermore, a threshold σ is set to determine the directional fluctuation. th For example, it can be set to 15° (or equivalent radian value), if the variability of a certain path direction meets the following condition: σ k ≤σ th If the path direction is consistent, it is considered to have good consistency and is determined to be a candidate hook edge segment with structural stability. Finally, the set of all paths that meet the direction consistency condition is output as follows:

[0134] Output the set of all paths marked as hook edge segments as hook edge contour features;

[0135] Specifically, each path It is composed of two-dimensional points connected in order of pixel coordinates, representing a local edge contour in the image. The overall contour features can be expressed as a set of paths and a pair of direction information for each path: Where: P k For the edge path pixel sequence, Θ k Let σ be the sequence of the main direction angles of the path. k Directional fluctuation;

[0136] The final output hook edge contour feature structure can be represented as: Each f k Includes: starting point coordinates (x) start ,y start ), endpoint coordinates (x end ,y end (mean direction angle) and line segment length L k ;

[0137] Specifically, the calculation of the hook's current spatial attitude parameters includes:

[0138] Extract the hook edge line segment set from the hook image, and represent each edge line segment as its start and end positions in the image coordinate system;

[0139] It should be noted that: an edge line segment refers to a pixel path in the image coordinate space that consists of continuous pixels, has a clear start and end point, and has a strong consistency in direction, representing the edge contour in the hook image that has a clear spatial orientation structure;

[0140] In implementation, for each edge path P k Extract the coordinates of its start and end points, the specific steps are as follows:

[0141] Path P k The first pixel (x) start ,y start ) is set as the starting point of this edge segment, and the last pixel (x) is... end ,y end Set as the endpoint of this edge segment;

[0142] (x) end ,y end ) and (x end ,y end ) combined into coordinate pairs (P) start ,P end Store the coordinate pair as input for subsequent direction and angle calculations;

[0143] It should be noted that the selection principle for the starting point and the ending point is "defined according to the pixel connection order". Even if there are spatial twists in the path, this rule ensures the consistency of the path direction and is consistent with the calculation order of the direction angle sequence.

[0144] The final output is a set of coordinates for the edge line segments: Among them, P start,k ,P end,k Path P k The coordinates of the starting and ending points;

[0145] Calculate the direction angle of all edge line segments, and extract direction aggregation clusters based on the direction angle set. The direction aggregation cluster is a subset of direction angles whose direction angle distribution density is greater than the aggregation judgment threshold.

[0146] Specifically, the calculation of the direction angle includes:

[0147] For each line segment (P) start,k ,P end,k Let the starting point be (x). s ,y s ), the endpoint is (x e ,y e );

[0148] Calculate the direction vector components:

[0149]

[0150] Calculate the unit direction vector:

[0151]

[0152] Among them, L k The length of the line segment is calculated as follows:

[0153]

[0154] Convert the unit direction vector to an angle value (principal direction angle):

[0155] θ k =arctan 2(y e -y s ,x e -x s );

[0156] Combining the angle values ​​of all line segments, we obtain the set of direction angles: Θ = {θ1, θ2, ..., θ n};

[0157] Specifically, the extracted direction aggregation cluster includes:

[0158] The set of direction angles Θ is segmented and statistically analyzed; specifically, the interval [0°, 360°) is divided into several interval segments according to a fixed angle step size (e.g., 10°).

[0159] Count the number of directional angles contained in each angle segment to obtain a histogram of directional angle distribution;

[0160] The directional angle distribution density of each directional segment is calculated and defined as:

[0161]

[0162] Where: n i Δθ represents the number of angles within the i-th direction segment; Δθ represents the width of the angle segment (i.e., the angle step size) (e.g., 10°).

[0163] Furthermore, a directional aggregation determination threshold D is set. th The parameters can be set based on sample statistics or experience (e.g., 5 directions / degree); from all direction segments, select the set of direction segments that meet the following conditions: Θ cluster ={θ k ∈Θ∣D(θ k )≥D th}, where D(θ) k () indicates the distribution density of the direction segment corresponding to the angle;

[0164] The final output is a cluster of directional aggregates: a subset of directional and angle aggregates. Corresponding edge segment set

[0165] Calculate the average directional angle of each directional angle in the directional cluster, and use this average directional angle as the tilt angle of the hook; calculate the offset directional angle between the centroid of the endpoint position of each directional angle in the directional cluster and the image center, and use this as the rotation angle of the hook.

[0166] It should be noted that: the tilt angle is used to describe the average orientation of the hook body's edge structure in the image; the rotation angle is used to describe the offset direction of the edge structure's center of gravity relative to the image center, reflecting whether the hook is off-center or misaligned.

[0167] Since the direction angle is periodic, directly calculating the arithmetic mean may lead to errors. Therefore, a vector average method is used:

[0168] For each direction angle θ i Convert to a unit direction vector:

[0169]

[0170] Calculate the average component of the direction vector:

[0171]

[0172] The inverse solution yields the average direction angle (tilt angle):

[0173]

[0174] To further obtain the spatial eccentricity trend of the hook structure in the image, it is necessary to calculate the offset direction of the overall centroid of the aggregated cluster line segment relative to the image center, which is defined as the rotation direction angle;

[0175] Specifically, the following operations are performed based on the start and end point coordinates of all edge segments in the directional aggregation cluster:

[0176] Calculate the center coordinates for each line segment:

[0177]

[0178] Find the centroid of all line segments:

[0179]

[0180] Obtain the geometric center coordinates (x0, y0) of the image, which is usually the center point of the image resolution, such as...

[0181] Calculate the directional angle (rotation angle) from the centroid to the image center:

[0182] θ rotate =arctan 2(y g -y0,x g -x0);

[0183] It should be noted that this rotation angle is used to reflect the offset trend of the cluster edge line segment relative to the image center, and is an important reference indicator for evaluating abnormal states such as hook misalignment or tilting.

[0184] The tilt angle and rotation angle are combined to form the spatial attitude parameters of the current hook.

[0185] Specifically, based on the tilt angle θ tilt With rotation angle θ rotate Combine them into a two-dimensional attitude vector Θ current =(θ tilt ,θ rotate This serves as the spatial attitude parameter of the current hook.

[0186] S103: Compare the current spatial attitude parameters with the preset hook standard attitude range to determine whether the current spatial attitude parameters belong to the hook standard attitude range.

[0187] It should be noted that the so-called "standard attitude range" refers to a two-dimensional angle tolerance range constructed based on the attitude data of the hook in a stable and secure state in historical image samples, which is used to determine the legality of the attitude parameters extracted from the current image.

[0188] In implementation, determining whether the pre-space attitude parameters belong to the hook standard attitude range includes:

[0189] Extract the hook tilt angle and hook rotation angle values ​​from the current spatial attitude parameters;

[0190] Obtain the tilt angle boundary range and rotation angle boundary range of the standard posture range of the hook. The tilt angle boundary range includes the upper boundary value and the lower boundary value of the tilt angle, and the rotation angle boundary range includes the upper boundary value and the lower boundary value of the rotation angle.

[0191] The hook tilt angle value and the tilt angle boundary range are compared respectively, and the hook rotation angle value and the rotation angle boundary range are compared respectively.

[0192] When the hook tilt angle value is within the tilt angle boundary range and the hook rotation angle value is within the rotation angle boundary range, it is determined that the current spatial attitude parameter belongs to the hook standard attitude range; otherwise, it is determined that it does not belong to the hook standard attitude range.

[0193] S104: When the current spatial attitude parameters do not belong to the standard attitude range of the hook, calculate the attitude adjustment vector and guide the hook correction actuator to perform the hook attitude correction operation based on the attitude adjustment vector;

[0194] In implementation, the calculation of the attitude adjustment vector includes:

[0195] Extract the hook tilt angle and hook rotation angle values ​​from the current spatial attitude parameters;

[0196] Obtain the corresponding tilt angle reference value and rotation angle reference value in the standard posture range of the hook, wherein the tilt angle reference value and rotation angle reference value are the center values ​​of the upper and lower boundaries of its standard posture range, respectively;

[0197] The difference between the current hook tilt angle value and the tilt angle reference value is calculated as the first attitude deviation component, and the difference between the current hook rotation angle value and the rotation angle reference value is calculated as the second attitude deviation component.

[0198] It should be noted that: the first attitude deviation component refers to the directional difference between the hook tilt angle and the standard tilt angle, reflecting the offset of the hook's main edge direction; the second attitude deviation component refers to the directional difference between the hook rotation direction angle and the standard rotation direction angle, reflecting the offset direction of the hook structure's center of gravity.

[0199] Specifically, let the current spatial attitude angle pair be Θ. current =(θ tilt ,θ rotate The standard attitude reference angle is... The calculation steps are as follows:

[0200] Calculate the difference between the current tilt angle value and the standard tilt angle reference value, and use it as the first attitude deviation component:

[0201]

[0202] Calculate the difference between the current rotation angle value and the standard rotation angle reference value, and use it as the second attitude deviation component:

[0203]

[0204] It should be noted that, since angles are periodic, in order to avoid errors in directional judgment, in actual engineering implementation, the difference should be normalized to a symmetrical interval, such as [-180°, +180°], to reflect the shortest rotation path of the directional difference.

[0205] The normalization process is as follows:

[0206] For any angle difference Δθ, perform the following standardization operation:

[0207]

[0208] Normalize the angle differences in the tilt and rotation directions to obtain the final effective attitude deviation components: First attitude deviation component: δ tilt Second attitude deviation component: δ rotate The final output is a two-dimensional attitude deviation structure: ΔΘ=(δ tilt ,δ rotate );

[0209] The first attitude deviation component and the second attitude deviation component are combined to form a two-dimensional vector, and the two-dimensional vector is output as an attitude adjustment vector to guide the hook attitude correction control operation.

[0210] Specifically, the two components are directly combined into a two-dimensional vector structure: The vector is used as the input parameter of the hook attitude correction control system. After receiving the vector, the control system will guide the actuator to adjust the angle around the horizontal and vertical axes of the hook according to the magnitude and sign of the angular offset in the two directions, so that the hook attitude gradually returns to the standard range.

[0211] Example 2

[0212] like Figure 2 As shown in the example, the parts not detailed in this embodiment are as shown in Example 1. This embodiment discloses a safety rope hook suspension status detection system for high-altitude operations, including:

[0213] The acquisition module 201 is used to map the sequence of center points of the acquired image to the coordinate space of the target surface, and determine the candidate set of stable regions based on the distribution density and time interval changes of the center points of the image in each grid area, and filter out the target hanging region according to the candidate set of stable regions.

[0214] The extraction module 202 is used to acquire the current hook image after obtaining the target suspension area, extract the hook edge contour features based on the hook image, and calculate the current spatial attitude parameters of the hook. The spatial attitude parameters include the tilt angle and rotation direction angle of the hook's main edge line segment.

[0215] The comparison module 203 is used to compare the current spatial attitude parameters with the preset hook standard attitude range to determine whether the current spatial attitude parameters belong to the hook standard attitude range.

[0216] The control module 204 is used to calculate the attitude adjustment vector when the current spatial attitude parameters do not belong to the standard attitude range of the hook, and guide the hook correction actuator to perform the correction operation on the hook attitude based on the attitude adjustment vector.

[0217] Example 3

[0218] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements any of the above-described methods for detecting the suspension status of a safety rope hook for high-altitude operations.

[0219] Since the electronic device described in this embodiment is the one used to implement the method for detecting the suspension status of a safety rope hook for high-altitude operations in this application embodiment, those skilled in the art can understand the specific implementation and various variations of the electronic device based on the method described in this application embodiment. Therefore, how the electronic device implements the method in this application embodiment will not be described in detail here. Any electronic device used by those skilled in the art to implement the method for detecting the suspension status of a safety rope hook for high-altitude operations in this application embodiment falls within the scope of protection of this application.

[0220] Example 4

[0221] This embodiment discloses a computer-readable storage medium, including a memory, a processor, and a computer program stored on the memory and running on the processor. When the processor executes the computer program, it implements any of the above-described methods for detecting the suspension status of a safety rope hook for high-altitude operations.

[0222] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters, weights, and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0223] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired or wireless network. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0224] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for detecting the suspension status of a safety rope hook used in high-altitude operations, characterized in that, include: The acquired image center point sequence is mapped to the target surface coordinate space. Based on the distribution density and time interval variation of the image center points in each grid region, a candidate set of stable regions is determined, and the target hanging region is selected according to the candidate set of stable regions; including: The target surface coordinate space is divided into multiple grid regions of fixed size, and each image center point in the image center point sequence is projected onto the target surface coordinate space and assigned to its corresponding grid region. For each grid region, count the number of image center points assigned to that grid region, and calculate the distribution density value obtained by dividing the number of image center points in that grid region by the image acquisition time; For each grid region, extract the image frame number sequence of the image center point distributed in the region, calculate the set of differences between adjacent frame numbers in the sequence, and perform a range-to-mean ratio operation on the set of differences to obtain the time interval variation coefficient of the region. Grid regions that simultaneously satisfy the conditions of having a distribution density value greater than or equal to the distribution density determination threshold and a time interval change coefficient less than or equal to the time interval change determination threshold will be included in the candidate set of stable regions. After acquiring the target suspension area, the current hook image is captured. Based on the hook image, the hook edge contour features are extracted, and the current spatial attitude parameters of the hook are calculated. The spatial attitude parameters include the tilt angle and rotation direction angle of the hook's main edge line segment. The current spatial attitude parameters are compared with the preset hook standard attitude range to determine whether the current spatial attitude parameters belong to the hook standard attitude range. When the current spatial attitude parameters do not belong to the standard attitude range of the hook, calculate the attitude adjustment vector, and guide the hook correction actuator to perform the correction operation on the hook attitude based on the attitude adjustment vector.

2. The method for detecting the suspension status of a safety rope hook for high-altitude operations according to claim 1, characterized in that, The step of filtering the target suspension region based on the candidate set of stability regions includes: In the image center point sequence, the corresponding frame number of each image center point in time sequence is extracted, and the grid region number to which each image center point belongs is recorded in sequence. The consistency of the grid region numbers to which multiple consecutive image center points belong is judged. If the grid region numbers to which multiple consecutive image center points belong are consistent with the grid region numbers in the same stability region candidate set, it is judged that the image center points stay continuously in the grid region. When the length of the frame number sequence in which the center point of an image stays continuously in a grid region of a certain stability region candidate set meets the preset dwell length determination condition, the grid region is marked as the target hanging region.

3. The method for detecting the suspension status of a safety rope hook for high-altitude operations according to claim 2, characterized in that, The extraction of hook edge contour features includes: Extract the set of pixels whose brightness changes are greater than the change in the global average brightness of the image from the hook image, and construct an image brightness change map; In the image brightness variation map, the set of continuous pixel paths is identified according to the connection relationship of pixel positions, and in each pixel path, the main direction angle sequence from the starting point to the ending point of the path is extracted; Calculate the standard deviation of the main direction angle sequence for each pixel path. If the standard deviation is less than the direction fluctuation threshold, then mark the path as a hook edge segment. Output the set of all paths marked as hook edge segments as hook edge contour features.

4. The method for detecting the suspension status of a safety rope hook for high-altitude operations according to claim 3, characterized in that, The calculation of the hook's current spatial attitude parameters includes: Extract the hook edge line segment set from the hook image, and represent each edge line segment as its start and end positions in the image coordinate system; Calculate the direction angle of all edge line segments, and extract direction aggregation clusters based on the direction angle set. The direction aggregation cluster is a subset of direction angles whose direction angle distribution density is greater than the aggregation judgment threshold. Calculate the average directional angle of each directional angle in the directional cluster, and use this average directional angle as the tilt angle of the hook; calculate the offset directional angle between the centroid of the endpoint position of each directional angle in the directional cluster and the image center, and use this as the rotation angle of the hook. The tilt angle and rotation angle are combined to form the spatial attitude parameters of the current hook.

5. The method for detecting the suspension status of a safety rope hook for high-altitude operations according to claim 4, characterized in that, The determination of whether the pre-space attitude parameters belong to the hook standard attitude range includes: Extract the hook tilt angle and hook rotation angle values ​​from the current spatial attitude parameters; Obtain the tilt angle boundary range and rotation angle boundary range of the standard posture range of the hook. The tilt angle boundary range includes the upper boundary value and the lower boundary value of the tilt angle, and the rotation angle boundary range includes the upper boundary value and the lower boundary value of the rotation angle. The hook tilt angle value and the tilt angle boundary range are compared respectively, and the hook rotation angle value and the rotation angle boundary range are compared respectively. When the hook tilt angle value is within the tilt angle boundary range and the hook rotation angle value is within the rotation angle boundary range, it is determined that the current spatial attitude parameter belongs to the hook standard attitude range; otherwise, it is determined that it does not belong to the hook standard attitude range.

6. The method for detecting the suspension state of a safety rope hook for high-altitude operations according to claim 5, wherein calculating the attitude adjustment vector includes: Extract the hook tilt angle and hook rotation angle values ​​from the current spatial attitude parameters; Obtain the corresponding tilt angle reference value and rotation angle reference value in the standard posture range of the hook, wherein the tilt angle reference value and rotation angle reference value are the center values ​​of the upper and lower boundaries of its standard posture range, respectively; The difference between the current hook tilt angle value and the tilt angle reference value is calculated as the first attitude deviation component, and the difference between the current hook rotation angle value and the rotation angle reference value is calculated as the second attitude deviation component. The first attitude deviation component and the second attitude deviation component are combined to form a two-dimensional vector, and this two-dimensional vector is output as an attitude adjustment vector to guide the hook attitude correction control operation.

7. A safety rope hook suspension status detection system for high-altitude operations, implemented based on the safety rope hook suspension status detection method for high-altitude operations according to any one of claims 1-6, characterized in that, include: The acquisition module is used to map the sequence of center points of the acquired images to the coordinate space of the target surface, and determine the candidate set of stable regions based on the distribution density and time interval changes of the image center points in each grid area, and filter out the target hanging region based on the candidate set of stable regions. The extraction module is used to acquire the current hook image after obtaining the target suspension area, extract the hook edge contour features based on the hook image, and calculate the current spatial attitude parameters of the hook. The spatial attitude parameters include the tilt angle and rotation direction angle of the hook's main edge line segment. The comparison module is used to compare the current spatial attitude parameters with the preset hook standard attitude range to determine whether the current spatial attitude parameters belong to the hook standard attitude range. The control module is used to calculate the attitude adjustment vector when the current spatial attitude parameters do not belong to the standard attitude range of the hook, and guide the hook correction actuator to perform the correction operation on the hook attitude based on the attitude adjustment vector.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the method for detecting the suspension status of a safety rope hook for high-altitude operations as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed, implements the method for detecting the suspension status of a safety rope hook for high-altitude operations as described in any one of claims 1 to 6.

Images