Crane hook locking state detection method, device, equipment and storage medium

Abstract

The application discloses a crane hook locking state detection method, device and equipment and a storage medium, and applies to the technical field of safety monitoring of hoisting equipment, and comprises the following steps: acquiring real-time working video data of a crane hook; extracting image frame data of each hook subcomponent under different visual angles from the real-time working video data; constructing a three-dimensional space scene containing all hook subcomponents based on all the image frame data; extracting each working section corresponding to each hook subcomponent from the three-dimensional space scene; obtaining real-time angular position information of the working section corresponding to the working relationship under each visual angle according to the working relationship between all hook subcomponents; and analyzing the real-time angular position information under each visual angle to obtain a locking state analysis result of the crane hook. The crane hook locking state detection method provided by the application can realize efficient, accurate and reliable crane hook locking state detection.

Description

Technical Field

[0001] This invention relates to the field of safety monitoring technology for lifting equipment, and in particular to a method, device, equipment, and storage medium for detecting the locked state of a crane hook. Background Technology

[0002] Cranes are widely used in construction, port loading and unloading, industrial manufacturing, and warehousing and logistics. During crane operation, the reliable locking of the hook directly affects the stability of the hoisted load and the safety of personnel, equipment, and the environment on site. Therefore, real-time and accurate detection of the hook's locking status is a crucial technical aspect for ensuring crane operation safety.

[0003] In existing technologies, operators need to climb onto an observation platform near the crane hook area to observe the hook's locking status in real time based on their own work experience. However, this visual inspection method is inefficient and poses safety risks due to the need for operators to frequently climb onto the observation platform. Furthermore, it is easily affected by factors such as the operator's experience, viewing angle deviation, and lighting conditions, leading to large errors in the judgment results and making it difficult to reliably and consistently provide the true locking status of the hook during crane operation. Summary of the Invention

[0004] This invention provides a method, apparatus, equipment, and storage medium for detecting the locking status of crane hooks, in order to solve the technical problems of low efficiency and low accuracy in detecting the locking status of crane hooks by manual visual inspection, and to achieve efficient, accurate, and reliable detection of the locking status of crane hooks.

[0005] To address the aforementioned technical problems, this invention provides a method for detecting the locked state of a crane hook, comprising: Acquire real-time video data of the crane hook's operation; Extract image frame data of each hook component from different perspectives from the real-time working video data; Based on all the image frame data, a three-dimensional spatial scene containing all the hook components is constructed; Extract the working segments corresponding to each hook component from the three-dimensional spatial scene; Based on the working relationship between all the hook components, the real-time angular position information of the working segment corresponding to the working relationship under each of the aforementioned viewpoints is obtained; The real-time angle position information under each viewpoint is analyzed to obtain the locking state analysis result of the crane hook.

[0006] As one preferred embodiment, constructing a three-dimensional spatial scene containing all the hook components based on all the image frame data includes: Image enhancement is performed on all the aforementioned image frame data to obtain enhanced image frame data; Target detection is performed on all the enhanced image frame data to obtain the spatial position information of each hook component; The three-dimensional spatial scene is constructed based on the spatial position information of each hook component.

[0007] As one preferred embodiment, constructing the three-dimensional spatial scene based on the spatial position information of each hook component includes: The spatial position information of each hook sub-component corresponding to each of the aforementioned viewpoints is analyzed to obtain the spatial association information of the sub-components; Based on the spatial association information of the sub-components, the three-dimensional spatial scene is constructed.

[0008] As one preferred embodiment, obtaining the real-time angular position information of the working segment corresponding to the working relationship from each viewing angle, based on the working relationship between all the hook components, includes: The working relationships between all the hook components are analyzed to determine the associated working segments from each perspective; The geometric relationships of each associated working segment are analyzed to obtain the real-time angular position information under the corresponding viewpoint.

[0009] As one preferred embodiment, the step of analyzing the geometric relationships of each of the associated working segments to obtain the real-time angular position information under the corresponding viewpoint includes: A first distance analysis is performed on the geometric relationships of all the associated working segments to obtain relative positional relationship information; A second pose analysis is performed on the geometric relationships of all the associated working segments to obtain angle deviation information; Based at least on the relative position relationship information and the angle deviation information, the real-time angle position information of the associated working segment under each viewpoint is determined.

[0010] As one preferred embodiment, the analysis of the real-time angular position information under each viewpoint to obtain the locking state analysis result of the crane hook includes: Based on the credibility information of all the image frame data, the real-time angle position information of each of the images is fused to obtain multi-view fusion information; The multi-view fusion information is analyzed to obtain the locking state analysis results of the crane hook.

[0011] As one preferred embodiment, the analysis of the multi-view fused information to obtain the locking state analysis result of the crane hook includes: The multi-view fusion information is analyzed to obtain a candidate set of hook states; Based on a preset threshold, the candidate set of hook states is analyzed to determine the analysis result of the locking state of the crane hook.

[0012] Another aspect of the present invention provides a crane hook locking state detection device, comprising: The data acquisition module is used to acquire real-time working video data of the crane hook; The image frame extraction module is used to extract image frame data of each hook component from different perspectives from the real-time working video data; A 3D scene construction module is used to construct a 3D spatial scene containing all the hook components based on all the image frame data. The working segment extraction module is used to extract each working segment corresponding to each hook component from the three-dimensional spatial scene; An angle position analysis module is used to obtain the real-time angle position information of the working segment corresponding to the working relationship under each viewing angle, based on the working relationship between all the hook components. The hook status analysis module is used to analyze the real-time angle position information under each viewpoint to obtain the locking status analysis result of the crane hook.

[0013] In another aspect, the present invention provides a crane hook locking state detection device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements a crane hook locking state detection method as described above.

[0014] In another aspect, the present invention provides a computer-readable storage medium storing a computer program, wherein when the device containing the computer-readable storage medium executes the computer program, it implements a crane hook locking state detection method as described above.

[0015] Compared to existing technologies, the beneficial effects of this invention are at least one of the following: This invention achieves remote, non-contact, automatic monitoring of the hook's locking state by acquiring real-time working video data of the crane hook and extracting image frame data of each hook component from different perspectives. This solves the personal safety risks caused by the need for operators to frequently climb onto the observation platform for visual inspection in existing technologies. Furthermore, this invention constructs a three-dimensional spatial scene containing all hook components based on all image frame data and extracts the corresponding working segments for each hook component from the three-dimensional spatial scene, achieving three-dimensional structured modeling of the hook locking mechanism. This solves the problem that existing visual inspection methods are susceptible to operator experience differences and subjective judgment biases. This invention addresses the issue of differential lighting conditions. By analyzing the working relationships between all hook components, it obtains the real-time angular position information of the working segment corresponding to each working relationship from each viewpoint. This analysis yields the locking state analysis result, enabling quantitative determination of the locking state from a spatial geometric perspective. This solves the problems of large judgment errors and difficulty in stably outputting the true locking state caused by changes in lighting conditions, single viewpoints, or obstructions in existing technologies. Furthermore, by constructing a three-dimensional spatial scene and extracting the real-time angular position information of the working segment, this invention achieves automatic, continuous, and quantitative detection of the hook locking state, solving the problems of low efficiency and inability to meet the needs of all-weather real-time monitoring in existing manual detection methods. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a crane hook locking state detection method in one embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of a crane hook component in one embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of a crane hook locking status detection device in one embodiment of the present invention; Figure 4 This is a structural block diagram of a crane hook locking status detection device according to one embodiment of the present invention.

[0017] Figure label: The components include: 11. Data acquisition module; 12. Image frame extraction module; 13. 3D scene construction module; 14. Working segment extraction module; 15. Angle and position analysis module; 16. Hook status analysis module; 21. Processor; 22. Memory. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] In the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," "third," etc., may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0020] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal communication between two components. The terms "vertical," "horizontal," "left," "right," "upper," "lower," and similar expressions used herein are for illustrative purposes only and do not indicate or imply that the device or component referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as limiting the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0021] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the invention. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0022] Cranes, as core equipment for material handling, are widely used in key sectors of the national economy such as construction, port loading and unloading, industrial manufacturing, warehousing and logistics, and large equipment installation. Their operational safety directly impacts production order, and the safety of personnel and property. In the continuous operation of a crane, the hook locking mechanism, as a core safety component preventing load detachment and ensuring stable attachment, is crucial. The reliability, proper positioning, and effectiveness of its locking state are key factors determining the stability of the entire lifting process, directly affecting the reliability of the load attachment, the personal safety of on-site personnel, the integrity of the crane itself, and the safety of the surrounding working environment. If the hook locking fails, is not fully closed, or forms a false locking state, it can easily lead to major safety accidents such as load detachment, falling objects, equipment overturning, and personal injury. Therefore, real-time, continuous, non-contact, high-precision, and interference-resistant intelligent detection of the hook locking state has become a key technical link in ensuring safe crane operation and preventing lifting safety risks, and is also a core technical requirement that urgently needs to be overcome in this field.

[0023] Existing hook locking status detection technologies generally rely on manual visual judgment, that is, professional operators climb to a high-altitude observation platform close to the crane hook and verify the hook locking status in real time based on their personal operating experience, visual observation and subjective judgment. This traditional detection method has several inherent flaws: First, operators need to frequently climb to heights, which not only results in low detection efficiency and high labor intensity, but also poses high-level personal safety risks such as falls and collisions, failing to meet safety production standards. Second, the detection results are highly dependent on the operator's experience, sense of responsibility, visual state, and subjective judgment, making them prone to human error and omissions, with extremely poor consistency and stability. Third, the detection process is easily affected by external environmental factors such as light intensity, backlight, shadows, rain, fog, and dust, and is also constrained by factors such as a single observation angle, hook swing, and partial obstruction, making it difficult to comprehensively and accurately reflect the actual spatial posture of the locking mechanism. This leads to significant errors in locking status determination and makes it impossible to stably, reliably, and accurately output the true locking status of the hook throughout the entire dynamic operation of the crane, failing to meet the technical requirements of intelligent, automated, and all-weather safety monitoring in modern crane operations.

[0024] One embodiment of the present invention provides a method for detecting the locked state of a crane hook. For details, please refer to [link / reference needed]. Figure 1 , Figure 1 The diagram shown is a flowchart illustrating a crane hook locking state detection method according to one embodiment of the present invention. The method includes steps S1 to S6: S1. Obtain real-time working video data of the crane hook; S2. Extract image frame data of each hook component from different perspectives from the real-time working video data; S3. Based on all the image frame data, construct a three-dimensional spatial scene containing all the hook components; S4. Extract the working segments corresponding to each hook component from the three-dimensional spatial scene; S5. Based on the working relationship between all the hook components, obtain the real-time angle position information of the working segment corresponding to the working relationship from each viewpoint; S6. Analyze the real-time angle position information under each of the aforementioned viewpoints to obtain the locking state analysis results of the crane hook.

[0025] The technical solution of this invention first acquires real-time video data during the operation of a crane hook, extracts corresponding image frame data of each sub-component of the hook from the multi-view video, obtains the spatial position of the sub-components through image enhancement and target detection, and constructs a three-dimensional spatial scene. Then, it extracts the working segments of each hook sub-component from the three-dimensional scene, analyzes the working relationship between each sub-component to obtain the real-time angle position information of the working segments under different perspectives, and finally fuses the angle position information based on the multi-view credibility and combines it with a preset threshold to complete the analysis and judgment, thereby outputting accurate and reliable locking state analysis results of the crane hook.

[0026] However, the hook locking state is a dynamic spatial attitude characteristic, and its true validity needs to be determined during continuous movement. A single static image cannot reflect the actual positioning state of the locking mechanism under swinging, rotating, and deflecting conditions, and is easily affected by momentary occlusion, reflection, and blurring interference, leading to misjudgment. At the same time, manual observation from height poses safety risks and subjective biases, and single-point sensors cannot cover changes in spatial attitude. Therefore, it is necessary to collect continuous, multi-view real-time video data to provide a complete, dynamic, and reliable original data source for subsequent 3D modeling, geometric relationship analysis, and locking state determination, so as to achieve non-contact, all-weather, and automated locking state detection.

[0027] Specifically, in step S1, high-definition image acquisition devices with synchronous acquisition capabilities are deployed at the fixed position and the mobile follow-up position in the crane hook operation area. These devices include a main-view industrial camera, at least one auxiliary-view auxiliary camera, and an optional side-view blind-filling camera. Each camera works collaboratively through a timestamp synchronization mechanism to continuously acquire video streams of the hook and its locking mechanism throughout the entire operation process at a preset frame rate. The video streams are then transmitted in real time to the processing unit via an image preprocessing interface to form real-time working video data that can be used for analysis.

[0028] The primary view video data is acquired by a primary view camera mounted on the crane trolley, facing the front of the hook locking mechanism. This data captures the complete frontal outline of the hook body, locking tongue, and opening area, providing dominant view data for core target detection and geometric relationship analysis. Secondary view video data is acquired by secondary view cameras positioned to the side of the primary view at a certain angle. This data acquires the lateral posture and spatial depth information of the locking mechanism, compensating for feature loss during swinging and deflection of the primary view, and providing complementary perspectives for 3D spatial construction. Blind spot video data is acquired by a follow-up blind spot camera that moves with the hook's lifting or boom swing. This data covers areas partially obscured by the suspended object, slings, and wire ropes, ensuring complete image acquisition of key areas of the locking mechanism even under complex working conditions. Global view video data is acquired by a panoramic camera mounted in the cab or on top of the boom. This data is used to locate the overall position of the hook and the working scene, providing scene reference information for multi-view target association and spatial calibration.

[0029] This step enables non-contact remote data acquisition, avoiding the safety hazards of operators climbing to observe, improving detection efficiency and operational continuity; it provides continuous dynamic video streams, capturing the entire attitude change of the locking mechanism from opening to closing, avoiding instantaneous interference from single-frame images; it supports multi-view synchronous acquisition, covering different orientations and attitudes of the hook, eliminating judgment errors caused by single-view occlusion and perspective distortion; and it enables real-time data transmission and processing, meeting the all-weather, real-time, and automated safety monitoring needs of cranes in dynamic operation.

[0030] Real-time video data is continuous temporal data; directly processing the entire video data would result in computational redundancy and insufficient real-time performance. The hook locking state is determined by the relative spatial attitudes of multiple sub-components. Single-view images suffer from occlusion, perspective distortion, and apparent misjudgment, failing to fully represent the true spatial relationships of the locking mechanism. Therefore, it is necessary to extract key image frames containing each hook sub-component from multi-view synchronous video, according to temporal and viewpoint dimensions. Target area data for 3D modeling and geometric relationship analysis should be separated and retained, providing standardized, multi-view, and component-specific basic inputs for subsequent 3D spatial scene construction and angle position calculation. This avoids invalid data interference and improves detection accuracy and system response speed.

[0031] Specifically, in step S2, the acquired multi-view synchronous real-time video data is time-aligned and frame-sampled, and image frames are extracted according to a preset frame rate or keyframe triggering strategy; foreground segmentation and component localization are performed on each frame based on the preset target region, background redundant regions are removed, and local image frame data corresponding to each hook sub-component under each viewpoint are extracted respectively; the extracted image frames are normalized, deduplicated, and quality-screened to form an image frame dataset classified and stored by viewpoint and component, ensuring that the features of each sub-component are complete, time-synchronized, and position-aligned under different viewpoints.

[0032] The hook assembly comprises all the functional components of the crane hook locking mechanism. For details, please refer to [link / reference]. Figure 2 , Figure 2 The diagram shows a structural schematic of a crane hook component in one embodiment of the present invention. In the diagram, 1 is the locking tongue, 2 is the hook body, including the hook tip, hook body, and inner boundary of the hook opening, 3 is the rotating shaft connector, and 4 is the crane body. The hook body, locking tongue, rotating shaft connector, and hook opening structure collectively determine whether the hook opening is completely closed and whether the locking is in place, and are the core analytical objects for determining the locking state.

[0033] This step extracts valid frames and component regions from the full video data, significantly reducing computational load and ensuring real-time detection; it focuses on the hook component region to eliminate scene interference and improve the accuracy of subsequent target detection and 3D modeling; it simultaneously retains sub-component images from different viewpoints to solve problems such as single-view occlusion, viewpoint deviation, and false locking; and it extracts components and viewpoints to provide structured input for subsequent 3D reconstruction and angle position analysis.

[0034] Single-view, two-dimensional image frames can only reflect the apparent position of the hook component, failing to characterize its true attitude, relative orientation, and depth relationship in three-dimensional space, making it difficult to distinguish between apparent locking and true locking. Furthermore, two-dimensional analysis is susceptible to the influence of viewpoint, occlusion, and swaying, leading to distortion in locking state determination. Therefore, it is necessary to fuse and reconstruct a unified and complete three-dimensional spatial scene based on multi-view image frame data, achieving a three-dimensional representation of the locking mechanism's spatial structure. This provides a true and accurate three-dimensional geometric foundation for subsequent segment extraction, angle and position calculation, and quantitative determination of the locking state.

[0035] Specifically, in step S3, firstly, adaptive image enhancement is performed on all image frame data to suppress interference from lighting, fog, motion blur, and metallic reflections, improving the edge and contour clarity of sub-components. Secondly, a target detection algorithm is used to perform frame-by-frame and view-by-view detection on the enhanced image frames, outputting the two-dimensional detection box and pixel coordinates of each hook sub-component, and converting them into spatial position information. This spatial position information includes at least the two-dimensional pixel coordinates of each sub-component under the corresponding viewpoint, the scale and orientation information of each sub-component, the calibrated camera projection coordinates and depth estimate, and the confidence level and occlusion degree of each sub-component. Next, the spatial positions of the same sub-component under each viewpoint are registered, fused, and error-corrected to analyze the spatial correlation information between sub-components. Finally, based on the geometric constraints of the participating viewpoints inside and outside the camera, and combining the spatial position information and spatial correlation information, a three-dimensional spatial scene containing all hook sub-components is reconstructed.

[0036] Image enhancement technology is an image processing method designed for complex imaging conditions during outdoor crane operations. It involves pixel-level optimization of the original image frame to improve the contrast of target features, suppress noise, and repair blurred areas. The specific processing flow includes: dynamic brightness suppression of bright areas and local brightening of low-light areas; linear contrast stretching and edge enhancement to highlight key contours such as latches, hooks, and shafts; removal of motion blur, rain and fog scattering, and dust interference to repair image distortion; and image normalization to unify resolution, brightness, and color gamut to ensure data consistency across multiple viewing angles.

[0037] Target detection technology is a visual perception technology that automatically locates and identifies hook components in complex backgrounds and outputs their category, location, and confidence level. Preferably, it is based on a lightweight convolutional neural network, which performs multi-scale feature extraction on the enhanced image to generate dense anchor boxes. The classification branch predicts the target category (hook body, latch, pivot, etc.) within the anchor box, and the regression branch corrects the anchor box coordinates, outputting detection results including target category, confidence level, and bounding box coordinates. After non-maximum suppression for deduplication, the precise location region of each sub-component in the image frame is obtained, achieving fast and high-precision component localization in complex scenes.

[0038] Spatial correlation information is obtained by jointly analyzing the spatial positions of different sub-components from different perspectives, resulting in information on the relative positions, constraints, and geometric matching between components. Specifically, it includes the corresponding matching relationship of the same sub-component across multiple perspectives, the relative distance, azimuth, and pitch angles between sub-components, the assembly constraints of the locking mechanism (such as the hinge relationship between the latch and the pivot, and the closed alignment relationship between the latch and the hook), and the geometric constraints of each sub-component, such as spatial collinearity, perpendicularity, and parallelism.

[0039] Preferably, the present invention provides an embodiment for constructing a three-dimensional scene. Specifically, firstly, let the third... v A perspective at any moment t The original image acquired is denoted as The image after adaptive image enhancement is denoted as The target detection result is recorded as , , This represents an image enhancement operator used to perform bright area suppression, dark area enhancement, contrast stretching, deblurring, and edge enhancement.

[0040] The enhanced image is directly used for target detection to address the challenges of accurately locating locking mechanisms in complex backgrounds, small target sizes, and under partial occlusion. The principle of the target detection technique is illustrated in the following formula: in, This represents the object detection operator.b i Indicates the first i Position parameters of candidate target boxes c i Indicates category label, p i Indicates the confidence level. N t Indicates time t The number of targets detected.

[0041] Secondly, the target detection results are transformed into the three-dimensional coordinates, attitude angles, and orientations of each sub-component in the world coordinate system, as well as the spatial assembly relationships, relative poses, gaps, and overlaps between the sub-components—that is, the spatial positional relationships of each sub-component. Finally, based on the spatial position and association information of the sub-components, a unified three-dimensional model environment containing all hook sub-components and their actual spatial attitudes is reconstructed.

[0042] The hook component is a complete mechanical structure, and its locking state is determined only by the key sections involved in the locking action, rather than the entire spatial range of the component. Directly using the complete component as the analysis unit would introduce invalid geometric data, leading to redundant calculations and decreased accuracy of key parameters such as angle, pose, and clearance. Therefore, it is necessary to extract the functional sections (i.e., working sections) from the constructed 3D spatial scene that directly participate in the locking engagement, bear the locking function, and determine the locking position. This allows for geometric focus, feature refinement, and precise quantification of locking state determination, providing the smallest, optimal, and most direct geometric calculation unit for subsequent angle position calculation and locking state analysis.

[0043] Specifically, in step S4, firstly, based on the mechanical design specifications of the crane hook locking mechanism, the key section ranges, geometric endpoints, reference axes, and boundary constraints of each sub-component involved in the locking action are predefined, establishing a priori rule base for extracting working sections. Secondly, based on the three-dimensional spatial scene, the three-dimensional model of each hook sub-component is segmented along the functional axis, distinguishing between non-working and working areas, and eliminating structural sections that do not participate in the locking engagement. Next, according to the priori rules, the starting endpoint, ending endpoint, central axis, and effective working surface of the working section are located on the segmented three-dimensional structure, and the three-dimensional geometric data of the corresponding section is extracted to form single, continuous, and computable working section data. Then, the extracted working sections are spatially aligned, attitude normalized, and geometrically corrected to unify the calculation benchmark of each working section, ensuring consistency and comparability of working sections extracted from different perspectives and at different times. Finally, based on the motion relationship of the locking mechanism, the length, position, and attitude of the working sections are verified to meet the mechanical engagement logic, abnormal sections are eliminated, and the final usable set of working sections for each sub-component is output.

[0044] The working section refers to the continuous three-dimensional geometric section on each hook component that directly participates in the closure of the hook opening, the reset and locking of the latch, the mechanical limit engagement, and undertakes the core locking function. It is the smallest functional analysis unit that reflects the degree of locking, the size of the gap, and whether the posture is correct. It includes at least the following: hook body working section (effective closing edge inside the hook opening, hook tip limiting section, hook opening upper edge contacting the latch), locking latch working section (latch free end sealing section, latch and hook opening alignment section, latch shaft drive section), shaft connector working section (shaft and latch hinged engagement section, shaft limiting rotation section), elastic reset component working section (latch abutting drive section, limiting rebound section), and limiting structure working section (latch maximum closed position limiting section, over-position blocking section).

[0045] The hook locking state is not determined independently by a single working section, but rather by the coordinated relationship between the working sections of each sub-component during the locking action. A single viewpoint can only acquire planar projection information and cannot directly reflect the actual spatial coordination relationship. Therefore, it is necessary to establish a correspondence between the viewpoint and the working sections based on the inherent mechanical coordination logic between each working section. Through geometric relationship analysis, the real-time angle and spatial position of each working section can be calculated, thereby transforming the three-dimensional structural information into quantifiable geometric feature parameters for determining the locking state, providing the core basis for the final locking state analysis.

[0046] Specifically, in step S5, firstly, based on the mechanical motion principle of the hook locking mechanism, the working relationships between each working segment, such as constraints, drives, limits, and blocking, are analyzed to identify the associated working segments that directly cooperate and interact during the locking action. Secondly, for each acquisition viewpoint, combined with the imaging projection relationship, the combination of associated working segments that can be clearly observed and stably calculated under that viewpoint is determined, excluding segments that are obscured or geometrically degraded. Next, relative distance analysis and pose angle analysis are performed on the associated working segments to calculate core geometric quantities such as gaps, offsets, relative angles, and attitude deviations between working segments. Finally, the calculation results of distance, pose, angle, gap, etc., are integrated and normalized according to the viewpoint dimension to form real-time angular position information of the corresponding working relationship under each viewpoint.

[0047] Preferably, the first distance analysis obtains the relative positional relationship information by calculating the Euclidean distance between each sub-component.

[0048] in,( x 1 , y 1 ) represents the coordinates of the bolt tip, ( x 2 , y 2The coordinates () represent the coordinates of the nearest point on the hook opening boundary. This distance parameter measures the relative positional relationship between the end of the locking tongue and the edge of the opening, and is an important geometric quantity for determining whether the locking is in place.

[0049] The second pose analysis obtains angular deviation information by calculating the relative pose deviations of each sub-component.

[0050] in, This indicates the current angle of the locking tongue relative to the hook. This indicates the standard locking angle. Angle deviation is used to reflect whether the bolt is in the correct return position.

[0051] Among them, real-time angular position information refers to the quantitative data such as spatial position, relative attitude, and geometric deviation calculated in real time based on the working relationship between related working segments under the corresponding viewpoint.

[0052] Real-time angle position information from a single perspective is susceptible to occlusion, viewpoint distortion, and transient interference, reflecting only a local apparent state and failing to directly represent the overall true locking state of the hook. This can easily lead to misjudgments such as false locking, partial locking, and apparent closure when actually not locked. Therefore, it is essential to perform reliability weighting, multi-source fusion, threshold judgment, and consistency verification on the angle position information calculated independently from multiple perspectives. This transforms scattered local geometric features into global, stable, and reliable locking state conclusions, ultimately outputting true locking state analysis results that can be directly used for safety monitoring and early warning.

[0053] Specifically, in step S6, firstly, based on the clarity, illumination quality, occlusion degree, feature integrity, and motion blur degree of each viewpoint image, the availability weight and confidence score corresponding to each frame image are calculated and assigned. Secondly, using availability information as weights, real-time angle, position, gap, deviation, and other data under each viewpoint are weighted and fused, low-confidence outliers are eliminated, single-viewpoint interference is suppressed, and multi-viewpoint fused information is generated. Next, the fused geometric parameters are classified and discriminated to generate a candidate set containing multiple possible states, and the confidence and matching degree of each candidate state are determined. The parameters in the state candidate set are compared with the locking qualification threshold, non-locking threshold, and suspected locking threshold, and threshold judgment and state constraints are performed. Finally, after time-series smoothing and state transition verification, the final analysis result is determined as follows: hook is currently locked, not locked, suspected locked, or locking abnormal.

[0054] Among them, the usability information of image frame data is a comprehensive evaluation index used to characterize the quality of image data at the corresponding viewpoint, whether it is suitable for angle position calculation, and whether it has the credibility of the decision. It includes: image sharpness score, contrast score, noise level; degree of interference such as motion blur, rain, fog, dust, and metallic reflection; visibility ratio, degree of occlusion, and integrity of hook components and working sections.

[0055] The hook state candidate set is a set of mutually exclusive or coexisting locking state candidates obtained after multi-view information fusion based on geometric parameters such as angle, position, gap, and attitude deviation. It is used to retain all reasonable state possibilities before the final threshold decision, avoiding premature judgment based on a single piece of evidence leading to misjudgment. It includes: reliably locked, not locked, partially locked, falsely locked, suspected locked, or locking anomaly.

[0056] Preferably, the present invention provides an embodiment for hook locking state analysis. First, multi-view fusion analysis is performed, which assigns a credibility weight to each viewpoint based on image quality and occlusion degree to solve the problem of inconsistent information reliability from different viewpoints.

[0057] in, exp(*) Represents an exponential function. Indicates the image quality score based on the viewpoint. Indicates the degree of occlusion. η 1 and η 2 Indicates the adjustment parameter. V Indicates the number of viewpoints. Weighting coefficients in multi-view fusion.

[0058] Then, the degree of divergence between viewpoints is penalized, thereby solving the problem of apparent locking misjudgment caused by perspective distortion and partial occlusion in a single viewpoint.

[0059] in, To improve the credibility of multi-perspective states, Var(*) This represents the variance of the closure scores for each perspective.

[0060] Next, the multi-view fusion information is smoothly accumulated to solve the problem of single-frame state fluctuation caused by interference such as hook swing, instantaneous reflection, and short-term occlusion.

[0061] in, λ ∈[0,1] represents the time-series smoothing coefficient.

[0062] in, = 1 This indicates that the truth has been locked. = 0 This indicates that the lock is not actually locked. = -1 This indicates a pending confirmation status. τ h and τ l These represent the high and low threshold values, respectively. τ h > τ l Finally, the true locking state of the crane hook is determined based on the threshold. Specifically, the true locking state of the hook must simultaneously meet two conditions: consistency of multiple perspectives and continuity of time, in order to confirm the true locking state rather than the apparent locking state.

[0063] Another embodiment of the present invention provides a crane hook locking status detection device; for details, please refer to [link to relevant documentation]. Figure 3 , Figure 3 The diagram shown illustrates the structure of a crane hook locking state detection device according to one embodiment of the present invention. The device includes: Data acquisition module 11 is used to acquire real-time working video data of the crane hook; Image frame extraction module 12 is used to extract image frame data of each hook component from different perspectives from the real-time working video data; The three-dimensional scene construction module 13 is used to construct a three-dimensional spatial scene containing all the hook components based on all the image frame data. The working segment extraction module 14 is used to extract each working segment corresponding to each hook component from the three-dimensional spatial scene; The angle position analysis module 15 is used to obtain the real-time angle position information of the working segment corresponding to the working relationship under each viewing angle based on the working relationship between all the hook components. The hook status analysis module 16 is used to analyze the real-time angle position information under each viewpoint to obtain the locking status analysis result of the crane hook.

[0064] Another embodiment of the present invention provides a crane hook locking status detection device. For details, please refer to [link to relevant documentation]. Figure 4 , Figure 4The diagram shows a structural block diagram of a crane hook locking state detection device according to one embodiment of the present invention. The crane hook locking state detection device provided in this embodiment includes a processor 21, a memory 22, and a computer program stored in the memory 22 and configured to be executed by the processor 21. When the processor 21 executes the computer program, it implements the steps described in the above embodiment of the crane hook locking state detection method, for example... Figure 1 Steps S1 to S6 as described above.

[0065] For example, the computer program may be divided into one or more modules, which are stored in the memory 22 and executed by the processor 21 to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the crane hook locking state detection device.

[0066] The crane hook locking status detection device may include, but is not limited to, a processor 21 and a memory 22. Those skilled in the art will understand that the schematic diagram is merely an example of a crane hook locking status detection device and does not constitute a limitation on such a device. It may include more or fewer components than illustrated, or combine certain components, or use different components. For example, the crane hook locking status detection device may also include input / output devices, network access devices, buses, etc.

[0067] The processor 21 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. The processor 21 is the control center of the crane hook locking status detection device, connecting all parts of the device via various interfaces and lines.

[0068] The memory 22 can be used to store the computer program and / or modules. The processor 21 implements various functions of the crane hook locking status detection device by running or executing the computer program and / or modules stored in the memory 22 and calling the data stored in the memory 22. The memory 22 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory 22 may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0069] The integrated module of the crane hook locking status detection device, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0070] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0071] Accordingly, embodiments of the present invention provide a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform steps in a crane hook locking state detection method as described in the above embodiments, for example... Figure 1 Steps S1 to S6 as described above.

[0072] Compared to existing technologies, the beneficial effects of this invention are at least one of the following: This invention achieves remote, non-contact, automatic monitoring of the hook's locking state by acquiring real-time working video data of the crane hook and extracting image frame data of each hook component from different perspectives. This solves the personal safety risks caused by the need for operators to frequently climb onto the observation platform for visual inspection in existing technologies. Furthermore, this invention constructs a three-dimensional spatial scene containing all hook components based on all image frame data and extracts the corresponding working segments for each hook component from the three-dimensional spatial scene, achieving three-dimensional structured modeling of the hook locking mechanism. This solves the problem that existing visual inspection methods are susceptible to operator experience differences and subjective judgment biases. This invention addresses the issue of differential lighting conditions. By analyzing the working relationships between all hook components, it obtains the real-time angular position information of the working segment corresponding to each working relationship from each viewpoint. This analysis yields the locking state analysis result, enabling quantitative determination of the locking state from a spatial geometric perspective. This solves the problems of large judgment errors and difficulty in stably outputting the true locking state caused by changes in lighting conditions, single viewpoints, or obstructions in existing technologies. Furthermore, by constructing a three-dimensional spatial scene and extracting the real-time angular position information of the working segment, this invention achieves automatic, continuous, and quantitative detection of the hook locking state, solving the problems of low efficiency and inability to meet the needs of all-weather real-time monitoring in existing manual detection methods.

[0073] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for detecting the locked state of a crane hook, characterized in that, include: Acquire real-time video data of the crane hook's operation; Extract image frame data of each hook component from different perspectives from the real-time working video data; Based on all the image frame data, a three-dimensional spatial scene containing all the hook components is constructed; Extract the working segments corresponding to each hook component from the three-dimensional spatial scene; Based on the working relationship between all the hook components, the real-time angular position information of the working segment corresponding to the working relationship under each of the aforementioned viewpoints is obtained; The step of obtaining the real-time angular position information of the working segment corresponding to the working relationship from each viewpoint based on the working relationship between all the hook components includes: The working relationships between all the hook components are analyzed to determine the associated working segments from each perspective; The geometric relationship of each of the associated working segments is analyzed to obtain the real-time angle position information under the corresponding viewpoint; The step of analyzing the geometric relationships of each of the associated working segments to obtain the real-time angular position information under the corresponding viewpoint includes: A first distance analysis is performed on the geometric relationships of all the associated working segments to obtain relative positional relationship information; A second pose analysis is performed on the geometric relationships of all the associated working segments to obtain angle deviation information; Based at least on the relative position relationship information and the angle deviation information, the real-time angle position information of the associated working segment under each viewpoint is determined; The real-time angle position information under each viewpoint is analyzed to obtain the locking state analysis result of the crane hook; The analysis of the real-time angle position information under each of the aforementioned viewpoints to obtain the locking state analysis result of the crane hook includes: Based on the credibility information of all the image frame data, the real-time angle position information of each of the images is fused to obtain multi-view fusion information; The multi-view fusion information is analyzed to obtain the locking state analysis results of the crane hook; The analysis of the multi-view fused information to obtain the locking state analysis result of the crane hook includes: The multi-view fusion information is analyzed to obtain a candidate set of hook states; Based on a preset threshold, the candidate set of hook states is analyzed to determine the analysis result of the locking state of the crane hook.

2. The method for detecting the locked state of a crane hook as described in claim 1, characterized in that, The construction of a three-dimensional spatial scene containing all the hook components based on all the image frame data includes: Image enhancement is performed on all the aforementioned image frame data to obtain enhanced image frame data; Target detection is performed on all the enhanced image frame data to obtain the spatial position information of each hook component; The three-dimensional spatial scene is constructed based on the spatial position information of each hook component.

3. The method for detecting the locked state of a crane hook as described in claim 2, characterized in that, The construction of the three-dimensional spatial scene based on the spatial position information of each hook component includes: The spatial position information of each hook sub-component corresponding to each of the aforementioned viewpoints is analyzed to obtain the spatial association information of the sub-components; Based on the spatial association information of the sub-components, the three-dimensional spatial scene is constructed.

4. A crane hook locking state detection device, applied to the crane hook locking state detection method according to claim 1, characterized in that, include: The data acquisition module is used to acquire real-time working video data of the crane hook; The image frame extraction module is used to extract image frame data of each hook component from different perspectives from the real-time working video data; A 3D scene construction module is used to construct a 3D spatial scene containing all the hook components based on all the image frame data. The working segment extraction module is used to extract each working segment corresponding to each hook component from the three-dimensional spatial scene; An angle position analysis module is used to obtain the real-time angle position information of the working segment corresponding to the working relationship under each viewing angle, based on the working relationship between all the hook components. The hook status analysis module is used to analyze the real-time angle position information under each viewpoint to obtain the locking status analysis result of the crane hook.

5. A crane hook locking status detection device, characterized in that, The system includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements a crane hook locking state detection method as described in any one of claims 1 to 3.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the device containing the computer-readable storage medium executes the computer program, it implements a crane hook locking state detection method as described in any one of claims 1 to 3.

Images