Anti-hoisting detection method and device, computer device, readable storage medium and program product

Abstract

The application relates to a hoisting prevention detection method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: acquiring monocular image data of a target area from multiple angles when a lifting appliance is lifted to a preset height; determining three-dimensional environment data of the target area based on the monocular image data; determining deformation data of the ground below the hoisted object, background point cloud data at a preset distance below the hoisted object and corner point position data of the hoisted object based on the three-dimensional environment data and a pre-trained visual large model; and determining a hoisting prevention detection result of a target vehicle based on the difference between the deformation data and initial deformation data, the difference between the background point cloud data and initial background point cloud data and the difference between the corner point position data and initial corner point position data. The method can achieve accurate and reliable hoisting prevention detection.

Description

Technical Field

[0001] This application relates to the field of intelligent industrial safety monitoring technology, and in particular to an anti-lifting detection method, device, computer equipment, computer-readable storage medium, and computer program product. Background Technology

[0002] With the development of intelligent industrial safety monitoring technology, anti-lifting identification methods based on three-dimensional perception have emerged. These methods mainly rely on point cloud data collected by lidar.

[0003] Traditional technologies typically employ lidar to scan the work environment and obtain 3D point clouds for accurate anti-lifting identification. However, lidar is prone to point cloud loss and quality degradation in extreme weather conditions such as rain and fog, and also suffers from high equipment costs and significant data processing delays. Therefore, there is an urgent need for a reliable anti-lifting identification method that can achieve reliable identification under complex working conditions while maintaining recognition accuracy, at a lower cost, and with higher robustness. Summary of the Invention

[0004] Therefore, it is necessary to provide a lifting prevention detection method, device, computer equipment, computer-readable storage medium, and computer program product that can achieve accurate and reliable detection to address the above-mentioned technical problems.

[0005] Firstly, this application provides a method for detecting anti-lifting, including:

[0006] When the lifting device is raised to a preset height, monocular image data of the target area is acquired from multiple angles. The target area includes the operating area of ​​the target vehicle, which is used to carry the object being lifted.

[0007] Based on the monocular image data, determine the three-dimensional environmental data of the target area;

[0008] Based on the three-dimensional environment data and the pre-trained visual large model, the deformation data of the ground below the hoisted object, the background point cloud data at a preset distance below the hoisted object, and the corner position data of the hoisted object are determined.

[0009] Acquire the initial deformation data of the ground below the hoisted object, the initial background point cloud data at a preset distance below the hoisted object, and the initial corner position data of the hoisted object when the hoisted object is stationary.

[0010] Based on the difference between the deformation data and the initial deformation data, the difference between the background point cloud data and the initial background point cloud data, and the difference between the corner position data and the initial corner position data, the anti-lifting detection result of the target vehicle is determined.

[0011] In one embodiment, determining the anti-lifting detection result of the target vehicle based on the difference between the deformation data and the initial deformation data, the difference between the background point cloud data and the initial background point cloud data, and the difference between the corner position data and the initial corner position data includes:

[0012] Based on the difference between the deformation data and the initial deformation data, a first intermediate detection result is determined;

[0013] The second intermediate detection result is determined based on the difference between the background point cloud data and the initial background point cloud data.

[0014] The third intermediate detection result is determined based on the difference between the corner position data and the initial corner position data;

[0015] Based on the joint probability distribution of the first intermediate detection result, the second intermediate detection result, and the third intermediate detection result, the anti-lifting detection result of the target vehicle is determined.

[0016] In one embodiment, determining the first intermediate detection result based on the difference between the deformation data and the initial deformation data includes:

[0017] If the difference between the deformation data and the initial deformation data is greater than a first preset threshold, the first intermediate detection result is determined to indicate that there is a risk of lifting.

[0018] In one embodiment, determining the second intermediate detection result based on the difference between the background point cloud data and the initial background point cloud data includes:

[0019] If the difference between the background point cloud data and the initial background point cloud data is less than a second preset threshold, the second intermediate detection result is determined to indicate that there is a risk of being lifted.

[0020] In one embodiment, determining the third intermediate detection result based on the difference between the corner position data and the initial corner position data includes:

[0021] If the difference between the corner position data and the initial corner position data does not match the preset height, the third intermediate detection result is determined to indicate a risk of lifting.

[0022] In one embodiment, determining the three-dimensional environment data of the target region based on the monocular image data includes:

[0023] Based on the monocular image data and the pre-trained monocular depth estimation model, the depth value of each pixel in the monocular image data is determined.

[0024] Based on the depth values ​​of each pixel and the camera's intrinsic parameter matrix, the three-dimensional point cloud data of the target area is determined through coordinate transformation.

[0025] Secondly, this application also provides an anti-lifting detection device, comprising:

[0026] The acquisition module is used to acquire monocular image data of a target area from multiple angles when the lifting device is raised to a preset height. The target area includes the operating area of ​​the target vehicle, which is used to carry the object being lifted.

[0027] The determination module is used to determine the three-dimensional environment data of the target area based on the monocular image data;

[0028] The determining module is also used to determine the deformation data of the ground below the hoisted object, the background point cloud data at a preset distance below the hoisted object, and the corner position data of the hoisted object based on the three-dimensional environment data and the pre-trained visual large model.

[0029] The acquisition module is used to acquire the initial deformation data of the ground below the hoisted object, the initial background point cloud data at a preset distance below the hoisted object, and the initial corner position data of the hoisted object when the hoisted object is in a stationary state.

[0030] The determining module is further configured to determine the anti-lifting detection result of the target vehicle based on the difference between the deformation data and the initial deformation data, the difference between the background point cloud data and the initial background point cloud data, and the difference between the corner position data and the initial corner position data.

[0031] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described in any of the above-mentioned embodiments.

[0032] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in any of the preceding claims.

[0033] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method described in any of the above claims.

[0034] The aforementioned anti-lifting detection method, device, computer equipment, computer-readable storage medium, and computer program product, by employing a monocular vision system combined with a large visual model to achieve three-dimensional environment reconstruction and multi-dimensional feature analysis, can avoid the problems of high cost, poor environmental adaptability, and complex multi-sensor synchronization in related technologies such as lidar systems. It achieves low-cost, highly robust, and high-precision anti-lifting identification in complex industrial scenarios. Furthermore, by combining three types of features—ground deformation, background point cloud changes, and corner distance changes—for comprehensive judgment, it improves the accuracy and robustness of detection and reduces erroneous operations caused by inaccurate judgment of a single feature. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 This is a diagram illustrating the application environment of the anti-lifting detection method in one embodiment;

[0037] Figure 2 This is a flowchart illustrating an anti-lifting detection method in one embodiment;

[0038] Figure 3 This is a flowchart illustrating step S150 in one embodiment;

[0039] Figure 4 This is a flowchart illustrating step S120 in one embodiment;

[0040] Figure 5 This is a structural block diagram of the anti-lifting detection device in one embodiment;

[0041] Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0043] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0044] The anti-lifting detection method provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or located on the cloud or other network servers. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, drones, low-altitude aircraft, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, projection devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. Head-mounted devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. Server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0045] In one exemplary embodiment, such as Figure 2 As shown, an anti-lifting detection method is provided, which is applied to... Figure 1 Taking terminal 102 as an example, the explanation includes the following steps S110 to S150. Wherein:

[0046] Step S110: When the lifting device is raised to a preset height, monocular image data of the target area is acquired from multiple angles. The target area includes the operating area of ​​the target vehicle, which is used to carry the object being lifted.

[0047] The object being lifted can be a container. The target vehicle can be a truck, train, etc. For example, multiple monocular cameras deployed on the crane's outriggers and lower crossbeam can acquire monocular image data of the target area containing the truck (or train) and container from different perspectives. Preset heights can include specific values ​​such as 30cm and 90cm to trigger detection at different stages. Specifically, the monocular camera on the outrigger can be used to capture the lifting features of the container's top surface. The monocular camera on the lower crossbeam can be used to capture the lifting features of the container's sides.

[0048] Step S120: Determine the three-dimensional environment data of the target area based on the monocular image data.

[0049] For example, the three-dimensional environmental point cloud of the target area can be reconstructed based on monocular image data by recovering the three-dimensional spatial geometric information from the two-dimensional image information.

[0050] Step S130: Based on the three-dimensional environment data and the pre-trained visual large model, determine the deformation data of the ground below the hoisted object, the background point cloud data at a preset distance below the hoisted object, and the corner position data of the hoisted object.

[0051] The large-scale visual segmentation model can be SAM (Segment Anything Model). For example, the 3D environment point cloud can be segmented based on the large-scale visual segmentation model to identify the ground, target vehicle and the object being hoisted, and to extract the feature data of the object being hoisted (including the deformation data of the ground under the target vehicle, the background point cloud data at a preset vertical distance under the object being hoisted, and the corner position data of the object being hoisted).

[0052] Step S140: Obtain the initial deformation data of the ground below the hoisted object, the initial background point cloud data at a preset distance below the hoisted object, and the initial corner position data of the hoisted object when the hoisted object is stationary.

[0053] For example, the initial values ​​of the above-mentioned feature data can be obtained before the hoisting operation begins and when the hoisted object is in a stationary state, so as to obtain initial deformation data, initial background point cloud data and initial corner position data for use as a reference.

[0054] Step S150: Based on the difference between the deformation data and the initial deformation data, the difference between the background point cloud data and the initial background point cloud data, and the difference between the corner position data and the initial corner position data, determine the anti-lifting detection result of the target vehicle.

[0055] For example, if the difference between the deformation data and the initial deformation data is less than a preset first threshold, the difference between the background point cloud data and the initial background point cloud data is less than a preset second threshold, and the difference between the corner position data and the initial corner position data matches a preset height, the anti-lifting detection result of the target vehicle can be determined as not being lifted.

[0056] The aforementioned anti-lifting detection method utilizes a monocular vision system combined with a large visual model to achieve 3D environment reconstruction and multi-dimensional feature analysis. This avoids the problems of high cost, poor environmental adaptability, and complex multi-sensor synchronization in related technologies such as lidar systems. It achieves low-cost, highly robust, and high-precision anti-lifting identification in complex industrial scenarios. Furthermore, by combining three types of features—ground deformation, background point cloud changes, and corner distance changes—for comprehensive judgment, the accuracy and robustness of detection are improved, and misoperations caused by inaccurate judgment of a single feature are reduced.

[0057] In some embodiments, such as Figure 3 As shown, step S150 above may include:

[0058] Step S501: Determine the first intermediate detection result based on the difference between the deformation data and the initial deformation data.

[0059] Step S502: Determine the second intermediate detection result based on the difference between the background point cloud data and the initial background point cloud data.

[0060] Step S503: Determine the third intermediate detection result based on the difference between the corner position data and the initial corner position data.

[0061] Step S504: Based on the joint probability distribution of the first intermediate detection result, the second intermediate detection result, and the third intermediate detection result, determine the anti-lifting detection result of the target vehicle.

[0062] Deformation data can refer to height or density data obtained from ground point cloud computing directly beneath the truck or train, and its trend reflects whether the ground has deformed due to the vehicle being lifted. Background point cloud data can refer to statistical characteristics calculated from point cloud distance values ​​in an area 20cm vertically below the lower edge of the container. Corner position data can refer to the three-dimensional spatial coordinates of the four corner points on the top surface of the container, used to calculate the change in corner distance.

[0063] For example, the first, second, and third intermediate detection results (i.e., their respective lifting state probability values) can be input into a pre-established joint probability distribution model. This model is calculated and statistically derived from a large amount of data collected on container lifting in various scenarios. It calculates the conditional probability of the lifting state when all three intermediate detection results occur simultaneously. If the probability value of the lifting state under this joint probability distribution is greater than 50%, the anti-lifting detection result for the target vehicle is determined to be a possibility of lifting, and the warning module is activated to stop the lifting operation. Once a lifting state is determined, if the probability value of the detection result from one of the methods is low, the system can prompt manual verification to check whether the sensor corresponding to that result is malfunctioning.

[0064] In one embodiment, step S501 above may include:

[0065] Step A: If the difference between the deformation data and the initial deformation data is greater than the first preset threshold, the first intermediate detection result is determined to indicate that there is a risk of lifting.

[0066] For example, the difference between the deformation data and the initial deformation data can be calculated in real time. If the difference (such as the percentage change in ground point cloud height) exceeds 5%, a lifting risk is determined, and a probability value indicating high risk is assigned to the first intermediate detection result; if the difference is within 5%, a probability value indicating low risk is assigned. This probability value can be obtained by statistically analyzing a large amount of lifting scenario data collected in advance.

[0067] Specifically, the ground and target vehicle can be segmented and identified based on a large visual segmentation model. Combined with the 3D environmental data reconstructed by the camera mounted on the crane's lower beam, the ground point cloud directly under the truck or train is defined as the monitoring area of ​​interest. When the lifting height of the spreader reaches 10-30cm, the changing trend of the ground point cloud within the monitoring area of ​​interest is compared in real time. If the increase in the ground point cloud is within 5%, it is determined that there is no risk of lifting the truck or train. If there is a large change in the ground point cloud, i.e., the change exceeds 5%, the first intermediate detection result is determined to be that there is a risk of lifting.

[0068] In one embodiment, step S502 above may include:

[0069] Step B: If the difference between the background point cloud data and the initial background point cloud data is less than the second preset threshold, the second intermediate detection result is determined to indicate that there is a risk of lifting.

[0070] For example, the difference between the background point cloud data and the initial background point cloud data can be calculated in real time. For instance, the proportion k of pixels whose point cloud distance values ​​change significantly in the region can be counted (e.g., the proportion of pixels whose difference is greater than a preset distance threshold). If the proportion k is lower than the preset proportion threshold, it is determined that there is a risk of lifting, and a probability value indicating high risk is assigned to the second intermediate detection result; otherwise, a probability value indicating low risk is assigned.

[0071] Specifically, the proportion of pixels (k) whose point cloud distance values ​​change significantly in the area 20cm vertically below the lower edge of the container (i.e., the percentage of pixels whose difference is greater than a preset distance threshold) can be counted. If the proportion k is less than 25%, the second intermediate detection result is determined to indicate that there is a risk of being lifted.

[0072] In one embodiment, step S503 above may include:

[0073] Step C: If the difference between the corner position data and the initial corner position data does not match the preset height, the third intermediate detection result is determined to indicate a risk of lifting.

[0074] For example, the difference between the corner position data and the initial corner position data can be calculated in real time to obtain the distance change of the container corner. This change is then compared with the lifting displacement of the spreader. If the two do not match (for example, the change is much smaller than the lifting displacement, indicating that the container may not have been lifted properly but was lifted along with the vehicle), a lifting risk is determined, and a probability value indicating high risk is assigned to the third intermediate detection result; if they match well, a probability value indicating low risk is assigned.

[0075] Specifically, the difference between the corner position data and the initial corner position data can be calculated in real time to obtain the distance change of the container corner; when the lifting displacement of the spreader reaches 30cm, this change is compared with the lifting displacement of the spreader; if the two do not match, the third intermediate detection result is determined to be that there is a risk of lifting.

[0076] In this embodiment, by combining three types of features—ground deformation, background point cloud changes, and corner distance changes—for independent analysis and multiple safety checks, deviations in intermediate results can be avoided when a sensor temporarily fails or a feature becomes abnormal due to interference from factors such as obstruction or changes in lighting (e.g., ground water affecting deformation analysis, or container side contamination affecting background point cloud judgment). Simultaneously, employing a joint probability distribution based on intermediate results enables dynamic weighting and fusion of detection results for different features, ensuring overall judgment reliability even when some features fail, thus improving fault tolerance and robustness. Furthermore, by comparing the ground deformation difference with a first preset threshold, rapid and quantitative judgment of changes in the contact state between vehicle tires and the ground can be achieved. When the ground rebounds significantly or recovers its deformation due to an accidental lifting of the vehicle, this physical change can be quickly captured, enabling direct and effective detection of the separation of the truck or train from the ground. By comparing the background point cloud difference with a second preset threshold, sensitive and indirect detection of the relative separation state between the container and the carrying vehicle can be achieved. When a container is lifted normally, the background area previously obscured by the vehicle gradually becomes visible, causing significant changes in the point cloud data at that location. Therefore, by quantifying this background visibility, it's possible to effectively detect whether the container and vehicle have separated as expected, thus determining if the vehicle is at risk of being lifted along with it. By comparing the corner point position difference with the spreader's preset height, accurate identification of abnormal container postures and trajectories can be achieved. If the actual lifting displacement of the container's corner points is inconsistent with the spreader's lifting height, it indicates that the container may not be simply lifted vertically, but rather has become hooked or pulled by the vehicle, resulting in abnormal movements such as tilting or jamming.

[0077] In one exemplary embodiment, such as Figure 4 As shown, step S120 above may include:

[0078] Step S1201: Based on the monocular image data and the pre-trained monocular depth estimation model, determine the depth value of each pixel in the monocular image data.

[0079] Step S1202: Based on the depth values ​​of each pixel and the camera's intrinsic parameter matrix, determine the 3D point cloud data of the target area through coordinate transformation.

[0080] For example, after completing the camera intrinsic parameter calibration, the depth value Z of each pixel in the image can be estimated based on the pre-trained monocular depth estimation model Metric3D V2; subsequently, the true physical coordinates X and Y corresponding to each pixel are calculated according to the following formulas, thereby reconstructing the 3D point cloud data:

[0081] (1)

[0082] (2)

[0083] Where u and v are the coordinate indices of the pixel in the pixel coordinate system, respectively; , These are the coordinate indices of the image center in the pixel coordinate system; , , respectively, are the horizontal and vertical focal lengths of the camera; Z is the depth value of each pixel estimated by the Metric3D V2 model.

[0084] In this embodiment, a pre-trained monocular depth estimation model is used to obtain depth information, and coordinate transformation is performed in combination with the camera intrinsic parameter matrix to generate a 3D point cloud. This eliminates the need for a complex binocular vision calibration and synchronization system, thereby reducing hardware costs and system complexity. Furthermore, by generating a high-density 3D point cloud that corresponds one-to-one with image pixels, more refined capture of details such as ground micro-deformation, container edge contours, and corner positions is achieved.

[0085] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0086] Based on the same inventive concept, this application also provides an anti-lifting detection device for implementing the anti-lifting detection method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more embodiments of the anti-lifting detection device provided below can be found in the limitations of the anti-lifting detection method described above, and will not be repeated here.

[0087] In one exemplary embodiment, such as Figure 5 As shown, an anti-lifting detection device 300 is provided, including: an acquisition module 301 and a determination module 302, wherein:

[0088] The acquisition module is used to acquire monocular image data of the target area from multiple angles when the spreader is raised to a preset height. The target area includes the operating area of ​​the target vehicle, which is used to carry the object being lifted.

[0089] The determination module is used to determine the three-dimensional environment data of the target area based on monocular image data;

[0090] The determination module is also used to determine the deformation data of the ground under the hoisted object, the background point cloud data at a preset distance under the hoisted object, and the corner position data of the hoisted object based on the 3D environment data and the pre-trained visual large model.

[0091] The acquisition module is used to acquire the initial deformation data of the ground below the hoisted object, the initial background point cloud data at a preset distance below the hoisted object, and the initial corner position data of the hoisted object when the hoisted object is stationary.

[0092] The determination module is also used to determine the anti-lifting detection result of the target vehicle based on the difference between the deformation data and the initial deformation data, the difference between the background point cloud data and the initial background point cloud data, and the difference between the corner position data and the initial corner position data.

[0093] In one embodiment, the determining module is further configured to:

[0094] The first intermediate detection result is determined based on the difference between the deformation data and the initial deformation data;

[0095] The second intermediate detection result is determined based on the difference between the background point cloud data and the initial background point cloud data;

[0096] The third intermediate detection result is determined based on the difference between the corner position data and the initial corner position data;

[0097] Based on the joint probability distribution of the first intermediate detection result, the second intermediate detection result, and the third intermediate detection result, the anti-lifting detection result of the target vehicle is determined.

[0098] In one embodiment, the determining module is further configured to:

[0099] If the difference between the deformation data and the initial deformation data is greater than the first preset threshold, the first intermediate detection result is determined to indicate that there is a risk of lifting.

[0100] In one embodiment, the determining module is further configured to:

[0101] If the difference between the background point cloud data and the initial background point cloud data is less than the second preset threshold, the second intermediate detection result is determined to indicate that there is a risk of lifting.

[0102] In one embodiment, the determining module is further configured to:

[0103] If the difference between the corner position data and the initial corner position data does not match the preset height, the third intermediate detection result is determined to indicate a risk of being lifted.

[0104] In one embodiment, the determining module is further configured to:

[0105] Based on monocular image data and a pre-trained monocular depth estimation model, the depth value of each pixel in the monocular image data is determined.

[0106] Based on the depth values ​​of each pixel and the camera's intrinsic parameter matrix, the 3D point cloud data of the target area is determined through coordinate transformation.

[0107] Each module in the aforementioned anti-lifting detection device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0108] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, input / output interfaces, a communication interface, a display unit, and an input device. The processor, memory, and input / output interfaces are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interfaces are used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When the computer program is executed by the processor, it implements an anti-lifting detection method. The display unit is used to form a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0109] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0110] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0111] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0112] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0113] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0114] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0115] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for detecting anti-lifting, characterized in that, The method includes: When the lifting device is raised to a preset height, monocular image data of the target area is acquired from multiple angles. The target area includes the operating area of ​​the target vehicle, which is used to carry the object being lifted. Based on the monocular image data, determine the three-dimensional environmental data of the target area; Based on the three-dimensional environment data and the pre-trained visual large model, the deformation data of the ground below the hoisted object, the background point cloud data at a preset distance below the hoisted object, and the corner position data of the hoisted object are determined. Acquire the initial deformation data of the ground below the hoisted object, the initial background point cloud data at a preset distance below the hoisted object, and the initial corner position data of the hoisted object when the hoisted object is stationary. Based on the difference between the deformation data and the initial deformation data, the difference between the background point cloud data and the initial background point cloud data, and the difference between the corner position data and the initial corner position data, the anti-lifting detection result of the target vehicle is determined.

2. The method according to claim 1, characterized in that, The determination of the anti-lifting detection result of the target vehicle based on the difference between the deformation data and the initial deformation data, the difference between the background point cloud data and the initial background point cloud data, and the difference between the corner position data and the initial corner position data includes: Based on the difference between the deformation data and the initial deformation data, a first intermediate detection result is determined; The second intermediate detection result is determined based on the difference between the background point cloud data and the initial background point cloud data. The third intermediate detection result is determined based on the difference between the corner position data and the initial corner position data; Based on the joint probability distribution of the first intermediate detection result, the second intermediate detection result, and the third intermediate detection result, the anti-lifting detection result of the target vehicle is determined.

3. The method according to claim 2, characterized in that, The step of determining the first intermediate detection result based on the difference between the deformation data and the initial deformation data includes: If the difference between the deformation data and the initial deformation data is greater than a first preset threshold, the first intermediate detection result is determined to indicate that there is a risk of lifting.

4. The method according to claim 2, characterized in that, The step of determining the second intermediate detection result based on the difference between the background point cloud data and the initial background point cloud data includes: If the difference between the background point cloud data and the initial background point cloud data is less than a second preset threshold, the second intermediate detection result is determined to indicate that there is a risk of being lifted.

5. The method according to claim 2, characterized in that, The determination of the third intermediate detection result based on the difference between the corner position data and the initial corner position data includes: If the difference between the corner position data and the initial corner position data does not match the preset height, the third intermediate detection result is determined to indicate a risk of lifting.

6. The method according to claim 1, characterized in that, The determination of the three-dimensional environment data of the target region based on the monocular image data includes: Based on the monocular image data and the pre-trained monocular depth estimation model, the depth value of each pixel in the monocular image data is determined. Based on the depth values ​​of each pixel and the camera's intrinsic parameter matrix, the three-dimensional point cloud data of the target area is determined through coordinate transformation.

7. A device for preventing lifting detection, characterized in that, The device includes: The acquisition module is used to acquire monocular image data of a target area from multiple angles when the lifting device is raised to a preset height. The target area includes the operating area of ​​the target vehicle, which is used to carry the object being lifted. The determination module is used to determine the three-dimensional environment data of the target area based on the monocular image data; The determining module is also used to determine the deformation data of the ground below the hoisted object, the background point cloud data at a preset distance below the hoisted object, and the corner position data of the hoisted object based on the three-dimensional environment data and the pre-trained visual large model. The acquisition module is used to acquire the initial deformation data of the ground below the hoisted object, the initial background point cloud data at a preset distance below the hoisted object, and the initial corner position data of the hoisted object when the hoisted object is in a stationary state. The determining module is further configured to determine the anti-lifting detection result of the target vehicle based on the difference between the deformation data and the initial deformation data, the difference between the background point cloud data and the initial background point cloud data, and the difference between the corner position data and the initial corner position data.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images