Abnormal posture detection method based on underground aerial passenger device
By extracting the key point information of the passenger object's skeleton on the downhole flight attendant device and identifying their limb status, the accuracy of passenger posture detection in underground operations is solved, and the safety and stability of transportation are improved.
Patent Information
- Application Number
- CN202411676094.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-21
- Publication Date
- 2025-03-28
AI Technical Summary
During the underground operation, it is difficult to accurately identify the passenger posture detection of the downhole flight attendant device, which affects the safety of passengers and the stability of transportation.
By obtaining the passenger area image of the downhole flight attendant device, extracting the key point information of the skeleton of the passenger object, obtaining the passenger posture, and identifying the limb status based on the posture to determine whether it is an abnormal posture.
It improves the accuracy of passenger attitude information acquisition and the accuracy of abnormal attitude recognition, enhances the safety and stability of the downhole flight attendant device, and reduces the probability of abnormal situations.
Smart Images
Figure CN119723656A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of mine safety and detection technology, and in particular to an abnormal posture detection method based on an underground overhead passenger device. Background Art
[0002] In the scenario of underground operations, underground workers can transport personnel through the underground overhead passenger device installed in the mine. During the operation of the underground overhead passenger device, the posture of the passengers has a certain degree of influence on the safety of the passengers during the transportation process.
[0003] Therefore, it is very important to detect the posture of passengers in underground overhead passenger devices. Summary of the invention
[0004] The purpose of this application is to solve one of the technical problems in the above technology at least to a certain extent.
[0005] In a first aspect, the present application provides a method for detecting abnormal postures based on an underground overhead passenger device, comprising: acquiring an overhead passenger device to be detected, and performing image acquisition on a passenger area of the overhead passenger device to obtain a candidate detection image of the passenger area; based on the candidate detection image, extracting skeleton key point information of a passenger object in the passenger area to obtain the passenger posture of the passenger object; according to the passenger posture, acquiring a limb state of the passenger object, and identifying whether the passenger object is an abnormal posture object according to the limb state.
[0006] The second aspect of the present application provides an abnormal posture detection device based on an underground overhead passenger device, including: an acquisition module, used to acquire the overhead passenger device to be detected, and to collect images of the passenger area of the overhead passenger device to obtain a candidate detection image of the passenger area; an extraction module, used to extract skeleton key point information of the passenger object in the passenger area based on the candidate detection image to obtain the passenger posture of the passenger object; and an identification module, used to acquire the limb state of the passenger object according to the passenger posture, and identify whether the passenger object is an abnormal posture object according to the limb state.
[0007] The third aspect of the present disclosure proposes an electronic device, comprising: a processor; a memory for storing executable instructions of the processor; wherein the processor is configured to execute instructions to implement the abnormal posture detection method based on an underground overhead passenger device as proposed in the first aspect above.
[0008] The fourth aspect of the present disclosure proposes a computer-readable storage medium. When the instructions in the computer-readable storage medium are executed by the processor of an electronic device, the electronic device can execute the abnormal posture detection method based on the underground overhead passenger device proposed in the first aspect above.
[0009] The abnormal posture detection method and device based on the underground overhead passenger device proposed in the present application obtains the candidate detection image of the passenger area of the overhead passenger device to extract the skeleton key point information of the passenger object, thereby obtaining the passenger posture of the passenger object, and obtaining the limb state of the passenger object according to the passenger posture, thereby identifying whether the passenger object is an abnormal posture object. In the present application, the passenger posture of the passenger object is obtained by extracting the skeleton key point information from the candidate detection image of the passenger area, thereby improving the accuracy of obtaining the posture information of the passenger object, identifying whether the passenger object is an abnormal posture object according to the limb state obtained from the passenger posture, and improving the accuracy of identifying abnormal posture objects, thereby improving the safety and stability of the underground overhead passenger device for the transportation of passengers, reducing the probability of abnormal situations in underground operations, and improving the safety of workers in underground operations.
[0010] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0011] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:
[0012] Figure 1 This is a flow chart of an abnormal posture detection method based on an underground overhead passenger device according to an embodiment of the present application;
[0013] Figure 2 This is a flow chart of an abnormal posture detection method based on an underground overhead passenger device according to another embodiment of the present application;
[0014] Figure 3 This is a schematic structural diagram of an abnormal posture detection device based on an underground overhead passenger device according to an embodiment of the present application;
[0015] Figure 4 A block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION
[0016] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.
[0017] The following describes an abnormal posture detection method based on an underground overhead passenger device in an embodiment of the present application with reference to the accompanying drawings.
[0018] Figure 1 FIG. 1 is a flow chart of an abnormal posture detection method based on an underground overhead passenger device according to an embodiment of the present application. Figure 1 As shown, the method includes:
[0019] S101, obtaining an overhead passenger device to be detected, and collecting images of a passenger area of the overhead passenger device to obtain a candidate detection image of the passenger area.
[0020] During daily operations in a mine, workers can be moved and transported through the space provided in the underground overhead passenger device for workers to sit on. In this scenario, the space can be determined as the passenger area of the underground overhead passenger device.
[0021] In the embodiment of the present application, there is a set limited range for the posture of the passengers in the passenger area. It can be understood that, compared with the passengers who are in the passenger area based on postures outside the limited range, the passengers who are in the passenger area based on postures within the limited range are safer during the operation of the underground overhead passenger device than the passengers who are in the passenger area based on postures outside the limited range.
[0022] It can be seen from this that during the daily operation of the underground overhead passenger device, it is necessary to detect the posture of the passenger objects in its passenger area to improve the safety of the passenger objects during the operation of the underground overhead passenger device.
[0023] Optionally, based on the image acquisition method in the related art, image acquisition can be performed on the passenger area of the underground overhead passenger device, and the acquired image for posture detection can be marked as a candidate detection image of the passenger area.
[0024] Among them, the acquisition of candidate detection images can be achieved through the image acquisition equipment configured on the underground overhead passenger device. The image acquisition equipment can be a high-resolution industrial camera or other equipment that can achieve high-definition image acquisition in an underground environment. No specific limitation is made here.
[0025] S102, based on the candidate detection image, extracting skeleton key point information of the passenger object in the passenger area to obtain the passenger posture of the passenger object.
[0026] In the embodiment of the present application, the posture information of the passenger object can be extracted from the candidate detection image to identify whether the passenger object is in an abnormal posture state.
[0027] Optionally, based on the skeleton key point information extraction algorithm in the related technology, the candidate detection image can be processed by the extraction algorithm, and then the skeleton key point information of the passenger object in the passenger area carried in the candidate detection image can be obtained according to the processing result of the extraction algorithm.
[0028] Furthermore, after obtaining the skeleton key point information of the passenger object, the skeleton key point information of the passenger object can be algorithmically processed based on the posture extraction algorithm in the relevant technology, and then the posture information of the passenger object is obtained according to the result of the algorithm processing, and it is determined as the passenger posture of the passenger object.
[0029] S103, acquiring the limb status of the passenger object according to the passenger's posture, and identifying whether the passenger object is an abnormal posture object according to the limb status.
[0030] In an embodiment of the present application, based on the extracted passenger posture, it can be identified whether the passenger object in the current passenger area is in an abnormal posture, wherein the posture that exceeds the limited range of resources preset by the underground overhead passenger device can be marked as an abnormal posture.
[0031] Optionally, after acquiring the passenger posture, the status of the limbs, head, torso and other parts of the current passenger object can be determined according to the specific information in the passenger posture, and the status can be marked as the limb status of the passenger object.
[0032] Furthermore, based on the limb state of the passenger object, the states of the passenger object's limbs, head, torso and other parts are obtained, so as to obtain the relationship between the passenger object's current posture and the preset posture limited range of the underground overhead passenger device, and then identify whether the passenger posture of the passenger object is in an abnormal posture based on the acquired relationship, and when the passenger posture is identified as being in an abnormal posture, determine that the passenger object is an abnormal posture object.
[0033] The abnormal posture detection method based on the underground overhead passenger device proposed in the present application obtains the candidate detection image of the passenger area of the overhead passenger device to extract the skeleton key point information of the passenger object, thereby obtaining the passenger posture of the passenger object, and obtaining the limb state of the passenger object according to the passenger posture, thereby identifying whether the passenger object is an abnormal posture object. In the present application, the passenger posture of the passenger object is obtained by extracting the skeleton key point information from the candidate detection image of the passenger area, thereby improving the accuracy of obtaining the posture information of the passenger object, identifying whether the passenger object is an abnormal posture object according to the limb state obtained from the passenger posture, and improving the accuracy of identifying abnormal posture objects, thereby improving the safety and stability of the underground overhead passenger device for the transportation of passengers, reducing the probability of abnormal situations in underground operations, and improving the safety of workers in underground operations.
[0034] In the above embodiment, the recognition of abnormal posture objects can also be combined with Figure 2 understand, Figure 2 FIG. 1 is a flow chart of an abnormal posture detection method based on an underground overhead passenger device according to another embodiment of the present application. Figure 2 As shown, the method includes:
[0035] S201, collecting images of the passenger area to obtain an initial detection image of the passenger area.
[0036] In the embodiment of the present application, the operating environment of the underground overhead passenger device may be in a low-light and dusty environment. In this scenario, image acquisition of the passenger area can be achieved by using an image acquisition device configured in the passenger area of the underground overhead passenger device, and the acquired image can be marked as an initial detection image of the passenger area.
[0037] The image acquisition device may be a high-resolution industrial camera or a device capable of acquiring high-definition images in an underground working environment, and no specific limitation is made here.
[0038] S202, performing image preprocessing on the initial detection image to obtain a preprocessed candidate detection image, wherein the image preprocessing at least includes image denoising processing, image enhancement processing and image normalization processing.
[0039] In an embodiment of the present application, the initial detection image can be preprocessed based on the image preprocessing algorithm in the relevant technology, wherein the initial detection image can be denoised by the Gaussian filtering method in the relevant technology to remove noise in the initial detection image, and the initial detection image can be enhanced by contrast enhancement and brightness adjustment, and the initial detection image can be standardized by unifying the size and format, thereby achieving the purpose of image preprocessing of the initial detection image, and further, the image after the image preprocessing of the initial detection image is determined as the candidate detection image to be detected.
[0040] S203, cropping the boundary box of the passenger object in the candidate detection image to obtain a target detection image of the passenger object.
[0041] In an embodiment of the present application, the bounding box of the passenger object in the candidate detection image can be cropped based on the bounding box cropping algorithm in the related art, wherein the bounding box of the passenger object in the candidate detection image can be determined, and then the candidate detection image can be cropped according to the determined bounding box, thereby obtaining an image including the passenger object and the bounding box of the passenger object.
[0042] Further, the cropped image is determined as a target detection image of the passenger object.
[0043] S204, obtaining a trained key point extraction model, and extracting skeleton key point information of the passenger object from the target detection image through the key point extraction model, wherein the skeleton key point information at least includes the key point coordinates of each skeleton key point and the connection relationship between each skeleton key point.
[0044] In the embodiment of the present application, the skeleton key points of the passenger object and the connection relationship between the skeleton key points can be obtained through the skeleton key point extraction method, so as to identify the posture of the passenger object in the passenger area.
[0045] Among them, a trained key point extraction model can be obtained, and the target detection image can be input into the model. Through the model, each skeleton key point of the passenger object carried in the target detection image and the connection relationship between each skeleton key point are extracted to obtain the corresponding output result.
[0046] Further, the information output by the model, including the skeleton key points and the connection relationship between the skeleton key points, is determined as the skeleton key point information of the passenger object.
[0047] S205 , acquiring the passenger posture of the passenger object according to the key point coordinates of each skeleton key point in the skeleton key point information and the connection relationship between the skeleton key points.
[0048] Optionally, a trained posture extraction model is obtained, and through the model capabilities of the posture extraction model, the relative positions of the key points between the skeleton key points are obtained according to the key point coordinates of each skeleton key point and the connection relationship between the skeleton key points, and the passenger posture of the passenger object is obtained according to the relative positions of the key points.
[0049] In the embodiment of the present application, the posture information of the passenger object in the passenger area can be determined by extracting the skeleton key point information of the passenger object.
[0050] Among them, based on the trained posture extraction model, based on the skeleton key points and the connection relationship between the skeleton key points, the positional relationship between the skeleton key points can be determined as the relative position of the key points between the skeleton key points, and then the posture information of the passenger object is determined according to the extracted relative position information, and marked as the passenger posture of the passenger object.
[0051] As an example, a corresponding key point extraction model can be constructed based on the open posture detection model (OpenPose) in the related art, and then the skeleton key point information of the passenger object can be obtained based on the trained key point extraction model, and the extracted skeleton key point information is expressed as K = {(x1, y1), (x2, y2), ..., (x n ,y n )}, where (x i ,y i ) represents the two-dimensional coordinate information of the skeleton key points of the passenger object, i takes values of 1, 2, ..., n, where n represents the total number of skeleton key points.
[0052] In this example, based on the coordinates of the extracted skeleton key points and the connection relationship between the skeleton key points, the skeleton key point information of the passenger object can be obtained.
[0053] Furthermore, by opening up the model capability of the posture detection model, the posture information of the passenger object is extracted based on the skeleton key point information of the passenger object, and the extracted posture information is determined as the passenger posture of the passenger object.
[0054] S206, obtaining the relative positions of the head, limbs and torso of the passenger object according to the passenger posture, and obtaining the limb status of the passenger object based on the relative positions of the limbs.
[0055] In the embodiment of the present application, abnormal posture can be identified through the relative position relationship between the head, limbs and torso of the passenger object, wherein the relative position between the head, limbs and torso of the passenger object can be marked as the relative position of the limbs of the passenger object.
[0056] Furthermore, the state information of the limbs of the passenger object is determined through the relative positions of the limbs of the passenger object as the limb state of the passenger object.
[0057] As an example, the passenger posture of the passenger object can be input into the PF2You Only Look Once Version 8 (YOLOv8) network in the related technology. Through the model capability of the YOLOv8 model, the relative positions of the head, limbs and torso of the passenger object can be extracted from the passenger posture of the passenger object, and then the limb state of the passenger object can be obtained.
[0058] S207: In response to the limb state indicating that the posture of the passenger object is abnormal, determining that the passenger object is identified as an object with abnormal posture.
[0059] Optionally, in response to the limb status indicating that the passenger object's head position is lower than the passenger object's waist position, and the passenger object's limbs are in a scattered state, it is determined that the passenger object is in a falling state, and the passenger object is identified as an abnormal posture object.
[0060] In the embodiment of the present application, there is a limited range of sitting postures in the passenger area of the underground overhead passenger device, so as to ensure the safe riding of passengers in the passenger area when the underground overhead passenger device is in operation.
[0061] Among them, the range of sitting postures limited by the underground overhead passenger device may include a sitting posture with both hands tightly holding the handrails and both feet placed in the specified position, and may also include other postures that can fix the passenger object in the safe sitting area provided in the passenger area of the underground passenger device, which is not specifically limited here.
[0062] In this scenario, when the limb status indicates that the passenger object's head is lower than the passenger object's waist and the passenger object's limbs are in a three-open state, it can be determined that the current passenger object may fall in the passenger area, and thus it can be determined that the passenger object in a falling state cannot be in a safe riding state in the passenger area.
[0063] Further, it can be determined that the passenger posture of the passenger object in the scene is an abnormal passenger posture, and then the passenger object in the abnormal passenger posture can be determined as an abnormal posture object in the passenger area.
[0064] Optionally, in response to the limb status indicating that the torso of the passenger object is standing upright on the floor of the passenger area and the head position of the passenger object is higher than the waist position of the passenger object, it is determined that the passenger object is in a standing state, and the passenger object is identified as an abnormal posture object.
[0065] In the embodiment of the present application, passengers are prohibited from standing and / or walking in the passenger area of the underground overhead passenger device. In this scenario, when a passenger object is identified as standing or walking, the passenger posture of the passenger object can be determined to be an abnormal passenger posture.
[0066] Optionally, when the limb status indicates that the torso of the passenger object is standing upright in the passenger area and the head position of the passenger object is higher than the waist position, it can be determined that the passenger object is in a standing state. In this scenario, it can be determined that the passenger posture of the passenger object in this scenario is an abnormal passenger posture, and the passenger object in the abnormal passenger posture is determined as an abnormal posture object in the passenger area.
[0067] As an example, a posture recognition model constructed based on a convolutional neural network (CNN network) in the related art can be used to identify whether the passenger posture is abnormal, which can be understood in combination with the following expression:
[0068] S = f(K) = arg max(CNN(K))
[0069] Among them, S is the posture state category, f(K) represents the output of the key point information through the convolutional neural network, predicting the most likely state category, and judging whether the passenger is in a standing, falling or other abnormal behavior state by analyzing the joint angles, limb proportions, etc.
[0070] S208: In response to identifying the passenger object as an abnormal posture object, performing posture monitoring on the abnormal posture object to obtain a duration of the abnormal posture of the abnormal posture object.
[0071] In an embodiment of the present application, the passenger object may make timely adjustments autonomously after an abnormal posture occurs, or may continue to be in an abnormal posture. In this scenario, the duration that the passenger object is in the abnormal posture can be monitored, and the monitored duration can be determined as the duration that the passenger object's abnormal posture continues as an abnormal posture object.
[0072] Optionally, when a passenger object is identified as an abnormal posture object, the posture of the abnormal posture object can be monitored by a preset image acquisition device in the passenger area, and the duration of the abnormal posture can be counted by a preset timing device to obtain the duration of the abnormal posture of the abnormal posture object.
[0073] S209: In response to the abnormal posture duration being greater than or equal to a preset abnormal posture duration threshold, an abnormal posture alarm is issued to the abnormal posture object.
[0074] Optionally, when the acquired abnormal posture duration is greater than or equal to a preset posture duration threshold, it can be determined that the abnormal posture object cannot safely sit in the passenger area provided by the underground overhead passenger device. In this scenario, in order to ensure the safety and stability of the operation of the underground overhead passenger device, it is necessary to issue an abnormal posture alarm for the abnormal posture object.
[0075] Among them, the abnormal posture object can be given an audible and visual alarm through the alarm device preset in the passenger area, and the corresponding alarm information can be broadcast to the abnormal posture object through the broadcasting equipment preset in the passenger area, so as to achieve the purpose of giving an abnormal posture alarm to the abnormal posture object.
[0076] As an example, the abnormal posture alarm can be understood by combining the following expression:
[0077]
[0078] In the above expression, A represents the alarm signal, T represents the duration of the abnormal posture of the abnormal posture object, and T threshold Indicates the preset abnormal posture duration threshold.
[0079] Optionally, a preset abnormal posture risk processing strategy is obtained, and abnormal recovery processing is performed on the abnormal posture objects in the passenger area based on the abnormal posture risk processing strategy.
[0080] In an embodiment of the present application, after receiving the abnormal posture warning information, the abnormal posture object may not be able to autonomously recover to the passenger posture within the safe range. In this scenario, the corresponding abnormal posture risk processing strategy can be used to adjust and restore the passenger posture of the abnormal posture object in the passenger area to achieve abnormal recovery processing of the abnormal posture object.
[0081] As an example, when a passenger object is in a falling state in the passenger area and cannot get up on his own, the corresponding abnormal posture risk handling strategy can be obtained to help the passenger get up and return to a safe sitting posture with both hands holding the handrails and both feet placed in the specified position, so as to achieve the purpose of abnormal recovery processing of the abnormal posture object in this example.
[0082] As another example, when a passenger object is in a standing state in the passenger area and does not autonomously recover to a safe sitting position with both hands holding the handrails and both feet placed in the specified position after receiving the abnormal posture alarm, the preset abnormal posture risk handling strategy can be used to forcibly help the abnormal posture object to recover to a safe sitting position, so as to achieve the purpose of abnormal recovery processing of the abnormal posture object in this example.
[0083] The abnormal posture detection method based on the underground overhead passenger device proposed in the present application obtains the candidate detection image of the passenger area of the overhead passenger device to extract the skeleton key point information of the passenger object, thereby obtaining the passenger posture of the passenger object, and obtaining the limb state of the passenger object according to the passenger posture, thereby identifying whether the passenger object is an abnormal posture object. In the present application, the passenger posture of the passenger object is obtained by extracting the skeleton key point information from the candidate detection image of the passenger area, thereby improving the accuracy of obtaining the posture information of the passenger object, identifying whether the passenger object is an abnormal posture object according to the limb state obtained from the passenger posture, and improving the accuracy of identifying abnormal posture objects, thereby improving the safety and stability of the underground overhead passenger device for the transportation of passengers, reducing the probability of abnormal situations in underground operations, and improving the safety of workers in underground operations.
[0084] The abnormal posture detection methods based on underground overhead passenger devices proposed in the above-mentioned embodiments correspond to each other. An embodiment of the present disclosure further proposes an abnormal posture detection device based on underground overhead passenger devices. Since the abnormal posture detection device based on underground overhead passenger devices proposed in the embodiment of the present disclosure corresponds to the abnormal posture detection methods based on underground overhead passenger devices proposed in the above-mentioned embodiments, the implementation method of the abnormal posture detection method based on underground overhead passenger devices is also applicable to the abnormal posture detection device based on underground overhead passenger devices proposed in the embodiment of the present disclosure, and will not be described in detail in the following embodiments.
[0085] Figure 3 FIG. 1 is a schematic diagram of a structure of an abnormal posture detection device based on an underground overhead passenger device according to an embodiment of the present disclosure. Figure 3 As shown, the abnormal posture detection device 300 based on the underground overhead passenger device includes an acquisition module 31, an extraction module 32 and a recognition module 33, wherein:
[0086] An acquisition module 31 is used to acquire an overhead passenger device to be detected, and to collect images of a passenger area of the overhead passenger device to obtain a candidate detection image of the passenger area;
[0087] An extraction module 32, for extracting skeleton key point information of a passenger object in the passenger area based on the candidate detection image to obtain a passenger posture of the passenger object;
[0088] The identification module 33 is used to obtain the limb status of the passenger object according to the passenger posture, and identify whether the passenger object is an abnormal posture object according to the limb status.
[0089] In the embodiment of the present application, the extraction module 32 is also used to: crop the bounding box of the passenger object in the candidate detection image to obtain a target detection image of the passenger object; obtain a trained key point extraction model, and extract skeleton key point information of the passenger object from the target detection image through the key point extraction model, wherein the skeleton key point information at least includes the key point coordinates of each skeleton key point and the connection relationship between each skeleton key point; obtain the passenger posture of the passenger object according to the key point coordinates of each skeleton key point in the skeleton key point information and the connection relationship between each skeleton key point.
[0090] In the embodiment of the present application, the extraction module 32 is also used to: obtain a trained posture extraction model, and through the model capabilities of the posture extraction model, obtain the relative positions of the key points between each skeleton key point according to the key point coordinates of each skeleton key point and the connection relationship between each skeleton key point; and obtain the passenger posture of the passenger object according to the relative position of the key points.
[0091] In the embodiment of the present application, the identification module 33 is also used to: obtain the relative positions of the head, limbs and torso of the passenger object according to the passenger posture, and obtain the limb status of the passenger object based on the relative limb positions; in response to the limb status indicating that the posture of the passenger object is abnormal, determine and identify the passenger object as an abnormal posture object.
[0092] In the embodiment of the present application, the identification module 33 is further used to: in response to the limb status indication that the head position of the passenger object is lower than the waist position of the passenger object, and the limbs of the passenger object are in a scattered state, determine that the passenger object is in a falling state, and determine that the passenger object is identified as an object with an abnormal posture; in response to the limb status indication that the torso of the passenger object is upright on the floor of the passenger area, and the head position of the passenger object is higher than the waist position of the passenger object, determine that the passenger object is in a standing state, and determine that the passenger object is identified as an object with an abnormal posture.
[0093] In an embodiment of the present application, the acquisition module 31 is also used to: acquire images of the passenger area to obtain an initial detection image of the passenger area; perform image preprocessing on the initial detection image to obtain a preprocessed candidate detection image, wherein the image preprocessing includes at least image denoising processing, image enhancement processing, and image standardization processing.
[0094] In an embodiment of the present application, the device also includes an exception handling module, which is used to: in response to identifying a passenger object as an abnormal posture object, perform posture monitoring on the abnormal posture object to obtain the duration of the abnormal posture of the abnormal posture object; in response to the duration of the abnormal posture being greater than or equal to a preset abnormal posture duration threshold, issue an abnormal posture alarm for the abnormal posture object.
[0095] In the embodiment of the present application, the exception handling module is further used to: obtain a preset abnormal posture risk handling strategy, and perform abnormal recovery processing on the abnormal posture objects in the passenger area based on the abnormal posture risk handling strategy.
[0096] The abnormal posture detection device based on the underground overhead passenger device proposed in the present application obtains the candidate detection image of the passenger area of the overhead passenger device to extract the skeleton key point information of the passenger object, thereby obtaining the passenger posture of the passenger object, and obtaining the limb state of the passenger object according to the passenger posture, thereby identifying whether the passenger object is an abnormal posture object. In the present application, the passenger posture of the passenger object is obtained by extracting the skeleton key point information from the candidate detection image of the passenger area, thereby improving the accuracy of obtaining the posture information of the passenger object, identifying whether the passenger object is an abnormal posture object according to the limb state obtained from the passenger posture, and improving the accuracy of identifying abnormal posture objects, thereby improving the safety and stability of the underground overhead passenger device for the transportation of passengers, reducing the probability of abnormal situations in underground operations, and improving the safety of workers in underground operations.
[0097] To achieve the above embodiments, the present application also provides an electronic device, a computer-readable storage medium and a computer program product.
[0098] Figure 4 A block diagram of an electronic device according to an embodiment of the present application is shown in FIG. Figure 4 As shown, the device 400 includes a memory 41, a processor 42, and a computer program stored in the memory 41 and executable on the processor 42. When the processor 41 executes the program instructions, the execution is realized. Figure 1 to Figure 2 The embodiment proposes an abnormal posture detection method based on an underground overhead passenger device.
[0099] In order to implement the above embodiments, the present application also provides a non-transitory computer-readable storage medium storing computer instructions, the computer instructions are used to enable the computer to execute Figure 1 to Figure 2 The embodiment proposes an abnormal posture detection method based on an underground overhead passenger device.
[0100] In order to implement the above embodiment, the present application also provides a computer program product. When the instruction processor in the computer program product executes, Figure 1 to Figure 2 The embodiment proposes an abnormal posture detection method based on an underground overhead passenger device.
[0101] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.
[0102] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.
[0103] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.
[0104] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute the instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.
[0105] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0106] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.
[0107] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.
[0108] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.
Claims
1. A method for detecting abnormal posture of an underground overhead passenger device, characterized in that: The method comprises: Acquire an overhead passenger device to be detected, and collect images of a passenger area of the overhead passenger device to obtain a candidate detection image of the passenger area; Based on the candidate detection image, extracting skeleton key point information of the passenger object in the passenger area to obtain the passenger posture of the passenger object; According to the passenger's posture, the limb state of the passenger object is acquired, and according to the limb state, it is identified whether the passenger object is an abnormal posture object.
2. The method according to claim 1, characterized in that The step of extracting skeleton key point information of a passenger object in the passenger area based on the candidate detection image to obtain a passenger posture of the passenger object includes: Cropping a bounding box of a passenger object in the candidate detection image to obtain a target detection image of the passenger object; Acquire a trained key point extraction model, and extract the skeleton key point information of the passenger object from the target detection image through the key point extraction model, wherein the skeleton key point information at least includes the key point coordinates of each skeleton key point and the connection relationship between each skeleton key point; The passenger posture of the passenger object is acquired according to the key point coordinates of each skeleton key point in the skeleton key point information and the connection relationship between the skeleton key points.
3. The method according to claim 2, characterized in that The acquiring the passenger posture of the passenger object according to the key point coordinates of each skeleton key point in the skeleton key point information and the connection relationship between the skeleton key points includes: Obtaining a trained posture extraction model, and obtaining the relative positions of the key points between the skeleton key points according to the key point coordinates of the skeleton key points and the connection relationship between the skeleton key points through the model capability of the posture extraction model; The passenger posture of the passenger object is obtained according to the relative positions of the key points.
4. The method according to claim 1, characterized in that: The acquiring the limb state of the passenger object according to the passenger posture, and identifying whether the passenger object is an abnormal posture object according to the limb state, comprises: According to the passenger posture, obtaining the relative positions of the head, limbs and torso of the passenger object, and obtaining the limb state of the passenger object based on the relative positions of the limbs; In response to the limb status indicating that the posture of the passenger object is abnormal, it is determined that the passenger object is identified as an object with abnormal posture.
5. The method according to claim 4, characterized in that In response to the limb state indicating that the posture of the passenger object is abnormal, determining that the passenger object is identified as an object with abnormal posture includes: In response to the limb status indication that the head position of the passenger object is lower than the waist position of the passenger object, and the limbs of the passenger object are in a spread-out state, determining that the passenger object is in a falling state, and determining to identify the passenger object as the abnormal posture object; In response to the limb status indicating that the torso of the passenger object is standing upright on the floor of the passenger area and the head position of the passenger object is higher than the waist position of the passenger object, it is determined that the passenger object is in a standing state, and the passenger object is identified as the abnormal posture object.
6. The method according to claim 1, characterized in that The step of acquiring the overhead passenger device to be detected and collecting images of the passenger area of the overhead passenger device to obtain a candidate detection image of the passenger area includes: Capturing an image of the passenger area to obtain an initial detection image of the passenger area; The initial detection image is subjected to image preprocessing to obtain the preprocessed candidate detection image, wherein the image preprocessing at least includes image denoising processing, image enhancement processing and image normalization processing.
7. The method according to any one of claims 1 to 6, characterized in that The method further comprises: In response to identifying the passenger object as the abnormal posture object, performing posture monitoring on the abnormal posture object to obtain a duration of the abnormal posture of the abnormal posture object; In response to the abnormal posture duration being greater than or equal to a preset abnormal posture duration threshold, an abnormal posture alarm is issued to the abnormal posture object.
8. The method according to claim 7, characterized in that After the abnormal posture object is given an abnormal posture alarm in response to the abnormal posture duration being greater than or equal to a preset abnormal posture duration threshold, the method further includes: A preset abnormal posture risk processing strategy is obtained, and based on the abnormal posture risk processing strategy, abnormal recovery processing is performed on the abnormal posture object in the passenger area.
9. An abnormal posture detection device based on an underground overhead passenger device, characterized in that: The device comprises: an acquisition module, used for acquiring an overhead passenger device to be detected, and performing image acquisition on a passenger area of the overhead passenger device to obtain a candidate detection image of the passenger area; An extraction module, configured to extract skeleton key point information of a passenger object in the passenger area based on the candidate detection image, so as to obtain a passenger posture of the passenger object; The identification module is used to obtain the limb status of the passenger object according to the passenger posture, and identify whether the passenger object is an abnormal posture object according to the limb status.
10. An electronic device, characterized in that: include: processor; a memory for storing executable instructions for the processor; The processor is configured to execute instructions to implement the method as claimed in any one of claims 1 to 8.