Oil and gas surface engineering personnel behavior monitoring method

Abstract

The invention provides an oil and gas surface engineering personnel behavior monitoring method. The method comprises the following steps: acquiring image information of a target area; inputting the image information into a target detection model, and detecting whether a target human body exists in the image information; under the condition that the target human body is detected, determining the work type of the target human body, and outputting a human body image with the work type; inputting the human body image into a dressing identification model, outputting a safety helmet identification result and a tool identification result under the condition that the operation type is non-aerial operation, determining whether the target human body dressing is abnormal or not, and determining whether the target human body dressing is abnormal or not according to the safety helmet identification result, the safety belt identification result and the tool identification result under the condition that the operation type is aerial operation. Determining whether the target human body dressing is abnormal; inputting the human body image into a behavior recognition model, extracting behavior characteristics of the human body image, and determining whether the behavior of the target human body is abnormal; dressing abnormity and behavior abnormity monitoring can be carried out according to operation types of operators of oil and gas surface engineering.

Description

Technical Field

[0001] The present application relates to the technical field of oil and gas surface engineering monitoring, and in particular to a method for monitoring the behavior of oil and gas surface engineering personnel. Background Art

[0002] In recent years, the construction of oil and gas surface projects has progressed rapidly. However, due to the large area, many cross-operations, and large number of personnel at the construction site, there are certain management difficulties and safety accidents are prone to occur. If it relies on manual supervision, it is inefficient and prone to loopholes. Therefore, the existing camera equipment at the construction site is used to support computer vision to automatically monitor and alarm unsafe behaviors at the construction site, such as not wearing safety clothes, not wearing safety helmets, smoking, playing with mobile phones, etc. Compared with traditional supervision methods, it can effectively discover problems, reduce the workload of supervisors, and improve supervision efficiency. However, there are usually aerial workers and non-aerial workers at the construction site. The existing personnel behavior monitoring method cannot effectively distinguish between aerial workers and non-aerial workers in the image. There are differences in the unsafe behaviors corresponding to aerial workers and non-aerial workers, resulting in inaccurate monitoring results. Summary of the invention

[0003] In response to the above-mentioned problems, the present application provides a method for monitoring the behavior of oil and gas ground engineering personnel, which solves the problem that the existing personnel behavior monitoring methods cannot effectively distinguish between aerial workers and non-aerial workers in the image, and there are differences in the unsafe behaviors corresponding to aerial workers and non-aerial workers, resulting in inaccurate monitoring results.

[0004] The present application provides a method for monitoring the behavior of oil and gas surface engineering personnel, including:

[0005] Acquire image information of the target area;

[0006] Inputting the image information into the target detection model to detect whether there is a target human body in the image information;

[0007] When a target human body is detected, determining the operation type of the target human body, and outputting a human body image with the operation type; wherein the operation type includes aerial operation and non-aerial operation;

[0008] Input a human image into a clothing recognition model, and when the operation type is non-aerial operation, extract the head features and body features of the human image, output the safety helmet recognition result and the work clothes recognition result, and determine whether the target human's clothing is abnormal according to the safety helmet recognition result and the work clothes recognition result; when the operation type is aerial operation, extract the head features, waist features, and body features of the human image, output the safety helmet recognition result, the safety belt recognition result, and the work clothes recognition result, and determine whether the target human's clothing is abnormal according to the safety helmet recognition result, the safety belt recognition result, and the work clothes recognition result;

[0009] A human body image is input into a behavior recognition model to extract behavior features of the human body image. When the operation type is non-aerial operation, the behavior features are matched with a first preset behavior feature set to determine whether the target human body has abnormal behavior. When the operation type is aerial operation, the behavior features are matched with a second preset behavior feature set to determine whether the target human body has abnormal behavior.

[0010] In some embodiments, the acquiring image information of the target area includes:

[0011] Obtain video data of the target area;

[0012] Based on the number of video frames of the video data, the video data is split and processed to obtain an image sequence;

[0013] Images are sequentially acquired from the image sequence as image information of the target area.

[0014] In some embodiments, including:

[0015] The video data is split and processed at preset intervals to obtain an image sequence.

[0016] In some embodiments, inputting the image information into the target detection model to detect whether the target human body exists in the image information includes:

[0017] Inputting the image information into a target detection model to extract feature information from the image information;

[0018] The feature information is matched with preset human features to determine whether a target human body exists in the image information.

[0019] In some embodiments, including:

[0020] When there are multiple target human bodies, anchor frame labeling is performed on all target human bodies;

[0021] According to the position information of the anchor frame, the anchor frame is segmented from left to right or from top to bottom to generate the corresponding human body image.

[0022] In some embodiments, when a target human body is detected, determining the operation type of the target human body and outputting a human body image with the operation type includes:

[0023] In the case where the target human body is detected, detecting whether there is a scaffold in the image information;

[0024] In the case where no scaffold is detected, determining that the operation type of the target human body is non-aerial operation;

[0025] When a scaffold is detected, the operation type of the target person is determined based on the relationship between the body detection frame of the target person and the scaffold detection frame of the scaffold; wherein, if the bottom of the body detection frame is located within the scaffold detection frame, the operation type of the target person is aerial operation, otherwise, the operation type of the target person is non-aerial operation.

[0026] In some embodiments, including:

[0027] When the bottom of the body detection frame is located within the scaffold detection frame, obtaining first coordinate information of any point on the bottom of the body detection frame and second coordinate position of any point on the bottom of the scaffold detection frame;

[0028] Calculating distance information in a longitudinal direction between the first coordinate information and the second coordinate information;

[0029] If the distance information is greater than a preset distance, the target human body is performing an aerial operation, otherwise the target human body is performing a non-aerial operation.

[0030] In some embodiments, including:

[0031] When the operation type is non-aerial operation, a head detection frame and a body detection frame are generated based on the key point coordinates of the human body image;

[0032] Generate a head image according to the head detection frame, and extract head features based on the head image;

[0033] A body image is generated according to the body detection frame, and body features are extracted based on the body image.

[0034] In some embodiments, including:

[0035] When the operation type is an aerial operation, a head detection frame, a waist detection frame, and a body detection frame are generated based on the key point coordinates of the human body image;

[0036] Generate a head image according to the head detection frame, and extract head features based on the head image;

[0037] generating a waist image according to the waist detection frame, and extracting waist features based on the waist image;

[0038] A body image is generated according to the body detection frame, and body features are extracted based on the body image.

[0039] In some embodiments, including:

[0040] Generate an upper body detection frame and a lower body detection frame based on the body detection frame according to the key point coordinates of the human body image;

[0041] generating an upper body image according to the upper body detection frame, and extracting upper body features based on the upper body image;

[0042] Output clothing recognition results based on upper body features;

[0043] generating a lower body image according to the lower body detection frame, and extracting lower body features based on the lower body image;

[0044] Output pants recognition results based on lower body features;

[0045] Output the work clothes recognition result according to the clothes recognition result and the pants recognition result.

[0046] The present application provides a method for monitoring the behavior of oil and gas surface engineering personnel, which obtains image information of a target area; inputs the image information into a target detection model to detect whether a target person exists in the image information; when the target person is detected, the operation type of the target person is determined, and a human image with the operation type is output; the human image is input into a clothing recognition model, and when the operation type is non-aerial operation, the head features and body features of the human image are extracted, and a helmet recognition result and a work clothes recognition result are output; based on the helmet recognition result and the work clothes recognition result, it is determined whether the target person's clothing is abnormal, and when the operation type is aerial operation, the human image is output. , extract the head features, waist features and body features of the human image, output the helmet recognition results, safety belt recognition results and work clothes recognition results, and determine whether the target human is dressed abnormally according to the helmet recognition results, safety belt recognition results and work clothes recognition results; input the human image into the behavior recognition model, extract the behavior features of the human image, and when the operation type is non-aerial operation, match the behavior features with the first preset behavior feature set to determine whether the target human has abnormal behavior; when the operation type is aerial operation, match the behavior features with the second preset behavior feature set to determine whether the target human has abnormal behavior. It can monitor the clothing and behavior abnormalities of the operators of oil and gas surface engineering according to the operation type, and improve the accuracy of abnormal monitoring of operators. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Hereinafter, the present application will be described in more detail based on embodiments and with reference to the accompanying drawings.

[0048] Figure 1 A schematic diagram of the implementation flow of a method for monitoring the behavior of oil and gas surface engineering personnel provided in an embodiment of the present application;

[0049] Figure 2 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0050] In the drawings, the same reference numerals are used for the same components, and the drawings are not drawn to scale. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0052] In the following description, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0053] If similar descriptions of "first\second\third" appear in the application documents, the following instructions will be added. In the following description, the terms "first\second\third" involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\third" can be interchanged in a specific order or sequence where permitted, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein.

[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0055] Based on the problems existing in the related art, the embodiment of the present application provides a method for monitoring the behavior of oil and gas surface engineering personnel, and the executor of the monitoring method can be an electronic device. The electronic device can be various types of terminals such as a heat preservation device, a laptop computer, a tablet computer, a desktop computer, a set-top box, a mobile device (for example, a mobile phone, a portable music player, a personal digital assistant, a dedicated messaging device, a portable gaming device), and can also be implemented as a server. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms.

[0056] In some embodiments, the functions implemented by the monitoring method provided in the embodiments of the present application can be implemented by a processor of an electronic device calling a program code, wherein the program code can be stored in a computer storage medium.

[0057] The present application embodiment provides a method for monitoring the behavior of oil and gas surface engineering personnel. Figure 1 A schematic diagram of the implementation flow of a method for monitoring the behavior of oil and gas surface engineering personnel provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, including:

[0058] S1: Acquire image information of the target area;

[0059] In some embodiments, the step S1 includes:

[0060] S11: Acquire video data of the target area;

[0061] S12: splitting the video data based on the number of video frames of the video data to obtain an image sequence;

[0062] S13: sequentially acquiring images from the image sequence as image information of the target area.

[0063] In the embodiment of the present application, the image information of the target area can be obtained by collecting video data of the target area through a video acquisition device, and the corresponding image sequence is obtained by splitting the video frames of the video data, and the images in the image sequence are used as the image information of the target area.

[0064] In some embodiments, including:

[0065] S121: splitting the video data at preset time intervals to obtain an image sequence.

[0066] In an embodiment of the present application, when splitting and processing video data, video frames corresponding to the video data can be collected at preset intervals to obtain image sequences corresponding to the video frames, which can reduce the amount of image information and reduce the amount of data processing.

[0067] S2: inputting the image information into the target detection model to detect whether there is a target human body in the image information;

[0068] In some embodiments, the step S2 comprises:

[0069] S21: inputting the image information into a target detection model to extract feature information from the image information;

[0070] S22: Match the feature information with preset human features to determine whether a target human body exists in the image information.

[0071] S3: when a target human body is detected, determining the operation type of the target human body, and outputting a human body image with the operation type; wherein the operation type includes aerial operation and non-aerial operation;

[0072] In some embodiments, including:

[0073] S301: When there are multiple target human bodies, annotate all target human bodies with anchor frames;

[0074] S302: Segment the anchor frame from left to right or from top to bottom in sequence according to the position information of the anchor frame to generate a corresponding human body image.

[0075] In the embodiment of the present application, the target human body is marked by an anchor frame, and the anchor frame is segmented in an orderly manner from left to right or from top to bottom according to the position information of the anchor frame.

[0076] In some embodiments, step S3 includes:

[0077] S31: When a target human body is detected, detecting whether there is a scaffold in the image information;

[0078] S32: when no scaffold is detected, determining that the operation type of the target human body is non-aerial operation;

[0079] S33: When a scaffold is detected, the operation type of the target person is determined based on the relationship between the body detection frame of the target person and the scaffold detection frame of the scaffold; wherein, if the bottom of the body detection frame is located within the scaffold detection frame, the operation type of the target person is aerial operation, otherwise, the operation type of the target person is non-aerial operation.

[0080] In the embodiment of the present application, when the target human body is detected, it is necessary to detect whether there is a scaffold in the image information. When there is no scaffold in the image information, it is determined that the target human body's job type is non-aerial work, thereby outputting a human body image with non-aerial work. When there is a scaffold in the image information, the target human body's job type is determined based on the relationship between the target human body's body detection frame and the scaffold detection frame of the scaffold. If the bottom of the body detection frame is located within the scaffold detection frame, the target human body's job type is aerial work. Otherwise, the target human body's job type is non-aerial work. The target human body's job type can be quickly determined.

[0081] In some embodiments, including:

[0082] S331: when the bottom of the body detection frame is located within the scaffold detection frame, obtaining first coordinate information of any point on the bottom of the body detection frame and a second coordinate position of any point on the bottom of the scaffold detection frame;

[0083] S332: Calculate distance information between the first coordinate information and the second coordinate information in the longitudinal direction;

[0084] S333: If the distance information is greater than a preset distance, the target human body is performing an aerial operation; otherwise, the target human body is performing a non-aerial operation.

[0085] In an embodiment of the present application, the distance information in the longitudinal direction is calculated by using the first coordinate information of any point on the bottom of the body detection frame and the second coordinate position of any point on the bottom of the scaffolding detection frame, and the distance information is compared with the preset distance. If the distance information is greater than the preset distance, the target human body is performing aerial work, otherwise the target human body is not performing aerial work. This can avoid erroneously judging the operation type of the human body target as aerial work when the human body target is located on the ground inside the scaffolding.

[0086] S4: Inputting the human image into the clothing recognition model, and when the operation type is non-aerial operation, extracting the head features and body features of the human image, outputting the helmet recognition result and the work clothes recognition result, and determining whether the target human's clothing is abnormal according to the helmet recognition result and the work clothes recognition result; and when the operation type is aerial operation, extracting the head features, waist features, and body features of the human image, and outputting the helmet recognition result, the safety belt recognition result, and the work clothes recognition result, and determining whether the target human's clothing is abnormal according to the helmet recognition result, the safety belt recognition result, and the work clothes recognition result;

[0087] In some embodiments, including:

[0088] S41: When the operation type is a non-aerial operation, a head detection frame and a body detection frame are generated based on the key point coordinates of the human body image;

[0089] S42: generating a head image according to the head detection frame, and extracting head features based on the head image;

[0090] S43: Generate a body image according to the body detection frame, and extract body features based on the body image.

[0091] In an embodiment of the present application, in the case of non-aerial operations, a head detection frame and a body detection frame are generated based on the key point coordinates of the human body image. Through the head detection frame and the body detection frame, a head image and a body image are generated based on the human body image. The head features can be extracted through the head image, and the body features can be extracted through the body image.

[0092] In some embodiments, including:

[0093] S401: When the operation type is an aerial operation, a head detection frame, a waist detection frame, and a body detection frame are generated based on the key point coordinates of the human body image;

[0094] S402: generating a head image according to the head detection frame, and extracting head features based on the head image;

[0095] S403: generating a waist image according to the waist detection frame, and extracting waist features based on the waist image;

[0096] S404: Generate a body image according to the body detection frame, and extract body features based on the body image.

[0097] In an embodiment of the present application, in the case of aerial work, a head detection frame, a waist detection frame and a body detection frame are generated based on the key point coordinates of the human body image. A head image, a waist image and a body image are generated based on the human body image through the head detection frame, the waist detection frame and the body detection frame. The head features can be extracted through the head image, the waist features can be extracted through the waist image, and the body features can be extracted through the body image.

[0098] In some embodiments, including:

[0099] S4001: Generate an upper body detection frame and a lower body detection frame based on the body detection frame according to the key point coordinates of the human body image;

[0100] S4002: generating an upper body image according to the upper body detection frame, and extracting upper body features based on the upper body image;

[0101] S4003: Outputting clothing recognition results according to upper body features;

[0102] S4004: generating a lower body image according to the lower body detection frame, and extracting lower body features based on the lower body image;

[0103] S4005: outputting pants recognition results according to the lower body features;

[0104] S4006: Outputting a work clothes recognition result according to the clothes recognition result and the pants recognition result.

[0105] In the embodiment of the present application, there are more than one type of work clothes for the workers. When the workers mix and match different types of work clothes, the existing clothing recognition model cannot effectively recognize them. Therefore, the present application divides the body features into upper body features and lower body features, outputs the clothes recognition results through the upper body features, and outputs the pants recognition results through the lower body features. When the clothes recognition results and the pants recognition results match, the work clothes recognition results are normal, otherwise the work clothes recognition results are abnormal. This can improve the accuracy of the work clothes recognition results.

[0106] S5: Input the human body image into the behavior recognition model, extract the behavior features of the human body image, and when the operation type is non-aerial operation, match the behavior features with a first preset behavior feature set to determine whether the target human body has abnormal behavior; when the operation type is aerial operation, match the behavior features with a second preset behavior feature set to determine whether the target human body has abnormal behavior.

[0107] In an embodiment of the present application, the first preset behavior feature set corresponding to non-aerial operations and the second preset behavior feature set corresponding to aerial operations may be completely different or partially the same. Abnormal behavior images are provided in the first preset behavior feature set and the second preset behavior feature set. The similarity between the behavior features and the abnormal behavior images in the first preset behavior feature set or the second preset behavior feature set is calculated. If the similarity is greater than a preset threshold, it indicates that the behavior features match the abnormal behavior image, indicating that the target human behavior is abnormal.

[0108] In summary, by acquiring image information of the target area; inputting the image information into the target detection model to detect whether there is a target human body in the image information; in the case of detecting the target human body, determining the job type of the target human body, and outputting a human body image with the job type; inputting the human body image into the clothing recognition model, and in the case of the job type being non-aerial work, extracting the head features and body features of the human body image, outputting the helmet recognition result and the work clothes recognition result, and determining whether the target human body is dressed abnormally based on the helmet recognition result and the work clothes recognition result, and in the case of the job type being aerial work, extracting the head features of the human body image. The human body image is input into the behavior recognition model to extract the behavior features of the human body image. When the operation type is non-aerial operation, the behavior features are matched with the first preset behavior feature set to determine whether the target human body has abnormal behavior. When the operation type is aerial operation, the behavior features are matched with the second preset behavior feature set to determine whether the target human body has abnormal behavior. It can monitor the abnormal clothing and behavior of the operators of oil and gas surface engineering according to the operation type, and improve the accuracy of abnormal monitoring of operators.

[0109] It should be noted that in the embodiment of the present application, if the above-mentioned monitoring method is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the embodiment of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM, Read Only Memory), a disk or an optical disk. In this way, the embodiment of the present application is not limited to any specific combination of hardware and software.

[0110] Accordingly, an embodiment of the present application provides a storage medium on which a computer program is stored, characterized in that when the computer program is executed by a processor, the steps in the monitoring method provided in the above embodiment are implemented.

[0111] An embodiment of the present application provides an electronic device; Figure 2 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, the electronic device 400 includes: a processor 401, at least one communication bus 402, a user interface 403, at least one external communication interface 404, and a memory 405. The communication bus 402 is configured to realize the connection and communication between these components. The user interface 403 may include a display screen, and the external communication interface 404 may include a standard wired interface and a wireless interface. The processor 401 is configured to execute the program of the monitoring method stored in the memory to implement the steps in the monitoring method provided in the above embodiment.

[0112] It should be noted here that the description of the above storage medium and electronic device embodiments is similar to the description of the above method embodiments, and has similar beneficial effects as the method embodiments. For technical details not disclosed in the storage medium and device embodiments of this application, please refer to the description of the method embodiments of this application for understanding.

[0113] It should be understood that "one embodiment" or "an embodiment" mentioned throughout the specification means that specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It should be understood that in various embodiments of the present application, the size of the sequence number of the above-mentioned processes does not mean the order of execution, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application. The above-mentioned sequence numbers of the embodiments of the present application are only for description and do not represent the advantages and disadvantages of the embodiments.

[0114] It should be noted that, in this article, the terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, object or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, object or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, object or device including the element.

[0115] The above is only an implementation method of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for monitoring the behavior of oil and gas surface engineering personnel, characterized in that: include: Acquire image information of the target area; Inputting the image information into the target detection model to detect whether there is a target human body in the image information; When a target human body is detected, determining the operation type of the target human body, and outputting a human body image with the operation type; wherein the operation type includes aerial operation and non-aerial operation; Input a human image into a clothing recognition model, and when the operation type is non-aerial operation, extract the head features and body features of the human image, output the safety helmet recognition result and the work clothes recognition result, and determine whether the target human's clothing is abnormal according to the safety helmet recognition result and the work clothes recognition result; when the operation type is aerial operation, extract the head features, waist features, and body features of the human image, output the safety helmet recognition result, the safety belt recognition result, and the work clothes recognition result, and determine whether the target human's clothing is abnormal according to the safety helmet recognition result, the safety belt recognition result, and the work clothes recognition result; A human body image is input into a behavior recognition model to extract behavior features of the human body image. When the operation type is non-aerial operation, the behavior features are matched with a first preset behavior feature set to determine whether the target human body has abnormal behavior. When the operation type is aerial operation, the behavior features are matched with a second preset behavior feature set to determine whether the target human body has abnormal behavior.

2. The method according to claim 1, characterized in that The step of acquiring image information of the target area includes: Obtain video data of the target area; Based on the number of video frames of the video data, the video data is split and processed to obtain an image sequence; Images are sequentially acquired from the image sequence as image information of the target area.

3. The method according to claim 2, characterized in that include: The video data is split and processed at preset intervals to obtain an image sequence.

4. The method according to claim 1, characterized in that: The step of inputting the image information into the target detection model to detect whether a target human body exists in the image information comprises: Inputting the image information into a target detection model to extract feature information from the image information; The feature information is matched with preset human features to determine whether a target human body exists in the image information.

5. The method according to claim 1, characterized in that include: When there are multiple target human bodies, anchor frame labeling is performed on all target human bodies; According to the position information of the anchor frame, the anchor frame is segmented from left to right or from top to bottom to generate the corresponding human body image.

6. The method according to claim 1, characterized in that The method of determining the operation type of the target human body when the target human body is detected and outputting a human body image with the operation type attached includes: In the case where the target human body is detected, detecting whether there is a scaffold in the image information; In the case where no scaffold is detected, determining that the operation type of the target human body is non-aerial operation; When a scaffold is detected, the operation type of the target person is determined based on the relationship between the body detection frame of the target person and the scaffold detection frame of the scaffold; wherein, if the bottom of the body detection frame is located within the scaffold detection frame, the operation type of the target person is aerial operation, otherwise, the operation type of the target person is non-aerial operation.

7. The method according to claim 6, characterized in that include: When the bottom of the body detection frame is located within the scaffold detection frame, obtaining first coordinate information of any point on the bottom of the body detection frame and a second coordinate position of any point on the bottom of the scaffold detection frame; Calculating distance information in a longitudinal direction between the first coordinate information and the second coordinate information; If the distance information is greater than a preset distance, the target human body is performing an aerial operation, otherwise the target human body is performing a non-aerial operation.

8. The method according to claim 1, characterized in that: include: When the operation type is non-aerial operation, a head detection frame and a body detection frame are generated based on the key point coordinates of the human body image; Generate a head image according to the head detection frame, and extract head features based on the head image; A body image is generated according to the body detection frame, and body features are extracted based on the body image.

9. The method according to claim 1, characterized in that: include: When the operation type is an aerial operation, a head detection frame, a waist detection frame, and a body detection frame are generated based on the key point coordinates of the human body image; Generate a head image according to the head detection frame, and extract head features based on the head image; generating a waist image according to the waist detection frame, and extracting waist features based on the waist image; A body image is generated according to the body detection frame, and body features are extracted based on the body image.

10. The method according to claim 8 or 9, characterized in that: include: Generate an upper body detection frame and a lower body detection frame based on the body detection frame according to the key point coordinates of the human body image; generating an upper body image according to the upper body detection frame, and extracting upper body features based on the upper body image; Output clothing recognition results based on upper body features; generating a lower body image according to the lower body detection frame, and extracting lower body features based on the lower body image; Output pants recognition results based on lower body features; Output the work clothes recognition result according to the clothes recognition result and the pants recognition result.

Images