WO2023098419A1 - 人体越界的检测方法、装置和计算机可读存储介质 - Google Patents

人体越界的检测方法、装置和计算机可读存储介质 Download PDF

Info

Publication number
WO2023098419A1
WO2023098419A1 PCT/CN2022/130621 CN2022130621W WO2023098419A1 WO 2023098419 A1 WO2023098419 A1 WO 2023098419A1 CN 2022130621 W CN2022130621 W CN 2022130621W WO 2023098419 A1 WO2023098419 A1 WO 2023098419A1
Authority
WO
WIPO (PCT)
Prior art keywords
boundary
human body
protection
camera
area
Prior art date
Application number
PCT/CN2022/130621
Other languages
English (en)
French (fr)
Inventor
关涛
梅君君
葛成伟
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2023098419A1 publication Critical patent/WO2023098419A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/96—Management of image or video recognition tasks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30196—Human being; Person
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30232—Surveillance

Definitions

  • the embodiments of the present application relate to, but are not limited to, the technical field of human behavior detection, and in particular, relate to a method, device, and computer-readable storage medium for detecting a human body crossing a boundary.
  • Embodiments of the present application provide a method, device, and computer-readable storage medium for detecting a human body crossing a boundary.
  • the embodiment of the present application provides a detection method for a human body crossing the boundary, including: determining the protection boundary of the protection area; generating the boundary condition corresponding to the protection boundary; calculating the deployment position of the camera according to the protection boundary; The deployment position of the camera, adjust the coverage of the target camera shooting field of view, so that the protection boundary falls into the coverage; read the human body image captured by the target camera, and extract the key points of the human body image ; Detecting the transgression of the human body in the protection area according to the transgressive condition and the key points of the human body.
  • the embodiment of the present application provides a human body transgression detection device, including: a memory, a processor, and a computer program stored on the memory and operable on the processor, when the processor executes the computer program Realize the detection method of the human body crossing the boundary as described in the first aspect.
  • an embodiment of the present application provides an electronic device, including: a memory, a processor, and a computer program stored on the memory and operable on the processor, and the processor implements the above-mentioned first step when executing the computer program.
  • the detection method for a human body crossing a boundary is one aspect, the detection method for a human body crossing a boundary.
  • the embodiment of the present application provides a computer-readable storage medium, the computer-readable storage medium stores a computer-executable program, and the computer-executable program is used to make the computer perform the above-mentioned first aspect.
  • Fig. 1 is a main flow chart of a detection method for a human body crossing a boundary provided by an embodiment of the present application
  • Fig. 2A is a schematic diagram of a closed polygonal protected area provided by an embodiment of the present application.
  • Figure 2B is a schematic diagram of a closed recessed protected area provided by an embodiment of the present application.
  • Fig. 2C is a schematic diagram of a protected area provided by an embodiment of the present application as an unenclosed area
  • Fig. 3 is a schematic diagram of the camera coverage height and extension length on the protection boundary provided by an embodiment of the present application
  • Fig. 4 is a human body key point detection effect diagram provided by an embodiment of the present application.
  • Fig. 5 is a sub-flow chart of a detection method for a human body crossing a boundary provided by an embodiment of the present application
  • FIG. 6 is a sub-flow chart of a detection method for a human body crossing a boundary provided by an embodiment of the present application
  • Fig. 7 is a schematic top view of a rectangular enclosed protection area provided by an embodiment of the present application.
  • Fig. 8 is a sub-flow chart of a detection method for a human body crossing a boundary provided by an embodiment of the present application
  • Fig. 9 is a schematic top view of a long linear non-enclosed protection area provided by an embodiment of the present application.
  • Fig. 10 is a top view schematic diagram of adding a camera in a long linear non-enclosed protection area provided by an embodiment of the present application;
  • Fig. 11 is a schematic structural diagram of a human body transgression detection device provided by an embodiment of the present application.
  • Fig. 12 is a schematic structural diagram of an electronic device provided by an embodiment of the present application.
  • multiple means more than two, greater than, less than, exceeding, etc. are understood as not including the original number, and above, below, within, etc. are understood as including the original number. If there is a description of "first”, “second”, etc., it is only for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features or implicitly indicating the indicated The sequence relationship of the technical characteristics.
  • the embodiment of the present application provides a detection method, device and computer-readable storage medium for the human body crossing the boundary.
  • the protection boundary is set according to the protection area, and the inner side and The outer sides belong to the protection side or the non-protection side respectively, and the cross-border conditions for the human body to enter the protection boundary are set.
  • the video and images collected by the camera are analyzed to detect the key points of the human body.
  • the application can flexibly set the cross-border conditions for the human body to enter the protection boundary, extract the key points of the human body from the human body image captured by the camera, and judge the cross-border situation of the human body in the protected area by combining the cross-border conditions and the key points of the human body.
  • the application can determine the deployment position of the camera according to the protection boundary.
  • Accurately deployed multiple cameras can accurately reflect the demarcated boundary and the relative position of the nearby human body in the image, avoiding inaccurate images due to inconsistent viewing angles and boundaries. Reflect the relative position of the boundary line and the human body. Adjust the coverage of the field of view of each camera so that the protection boundary falls within the coverage, so as to determine the coverage area of the camera, allowing the designer to determine the position of the camera more accurately. For complex or large protection areas, you can visually see the coverage of cameras and the overlapping of each camera, which can effectively reduce the workload of purely manual calculation of coverage, and at the same time reduce the error of manually specifying positions, thereby reducing the number of cameras Time cost, communication cost, and financial cost caused by rework caused by incorrect deployment, additional cameras, structural parts, etc.
  • FIG. 1 is a flow chart of a method for detecting a human body crossing a boundary provided by an embodiment of the present application.
  • the detection method of human body crossing includes but not limited to the following steps:
  • Step 101 determining the protection boundary of the protection area
  • Step 102 generating a boundary-crossing condition corresponding to the protection boundary
  • Step 103 calculating the deployment position of the camera according to the protection boundary
  • Step 104 adjusting the coverage of the field of view of the target camera according to the deployment position of the camera, so that the protection boundary falls within the coverage;
  • Step 105 read the human body image captured by the camera, and extract the human body key points of the human body image;
  • Step 106 detecting the crossing of the human body in the protection area according to the crossing condition and the key points of the human body.
  • the protection area includes a closed area and a non-enclosed area.
  • the closed area can be a closed polygon or a closed concave shape, and the protection boundary fits the area that needs to be protected.
  • the protected area; the non-enclosed area is shown in Figure 2C, such as a straight line or a broken line, which can define a certain side of the protected boundary as the protected area, which is conducive to the detection of cross-border at places such as entrances and exits. Therefore, the protection boundary can be set according to the shape and needs of the actual site for the protection area.
  • the out-of-boundary condition may include a human body part out-of-boundary condition and a time period out-of-boundary condition.
  • the human body part out of bounds condition settings can be made to restrict specific human body parts according to the needs of the actual protection scene.
  • the human body part out of bounds condition can be set to be out of bounds for hands and not for feet.
  • it can also be set by default that all parts of the human body cannot cross the boundary.
  • the time period crossing condition it is also possible to set a limit to a specific time period according to the needs of the actual protection scene.
  • the time period crossing condition can be set to cross the boundary during the day and not to cross the boundary at night.
  • the deployment position of the camera may be determined by the endpoint position of the protection boundary and the coverage of the camera's field of view.
  • the setting method of a single camera looking sideways or squinting will cause false alarms that people do not enter the protection area but are judged to have crossed the boundary, and false alarms that the human body actually enters the protection area but is judged to have not entered.
  • the present application can accurately deploy the cameras at appropriate positions on the protection boundary, so that the images taken by each camera can accurately reflect the demarcated boundaries and the relative positions of the nearby human bodies, avoiding inaccurate images due to inconsistent viewing angles and boundaries. It reflects the relative position of the boundary line and the human body, and overcomes the problems of false positives and false negatives of the human body crossing the boundary existing in the prior art.
  • each camera has its own coverage of the shooting field of view
  • the coverage of the camera's shooting field of view can be adjusted by adjusting the camera's field of view parameters, hanging height, orientation angle, and optical axis direction.
  • the coverage of the field of view of the corresponding camera needs to be adjusted so that the entire protection boundary can fall within the coverage.
  • the boundary line within the coverage area of the camera is superimposed on the picture captured by the camera. Observe whether the boundaries of adjacent cameras match. If there is a deviation, adjust the boundaries. Eliminate the incompatibility of the inner boundaries of adjacent camera screens caused by camera installation errors.
  • adjusting the coverage of the camera's shooting field of view also includes adjusting the coverage height and extension of each camera.
  • the coverage height is the height of the human body captured by the camera at the protection boundary.
  • the coverage height between the fields of view of any two adjacent cameras is generally set to be greater than the height of the human body to ensure that the entire human body on the boundary line can be covered by the camera.
  • the external extension is the horizontal extension length of the camera when the human body enters the protection boundary.
  • the extension will extend to the non-protection side of the protection boundary.
  • the external extension of the camera is generally set to about one meter to ensure that the camera can capture the entire human body entering the protection.
  • the human body image is captured by a camera, and the key points of the human body can be extracted from the human body image using a deep learning method.
  • a deep learning method For example, as shown in Figure 4, you can use deep learning algorithms such as openpose to detect images captured by cameras and extract key points of the human body.
  • the use of deep learning methods to identify human bodies can well adapt to environmental changes such as lighting, and has a strong adaptability to environmental changes such as lighting and ground reflections.
  • this application can flexibly set the transgression conditions of the human body entering the protection boundary, extract the key points of the human body from the human body image captured by the camera, and judge the transgression of the human body in the protection area by combining the transgression conditions and the key points of the human body.
  • the key points of the human body may include the head, torso, upper limbs, lower limbs, hands, feet, etc. of a person.
  • the cross-border condition to prohibit all parts of the human body from entering at night. Therefore, when any key point of the human body is determined to enter the rectangular protection boundary at night, it is determined that the human body has entered within the protected area of plant equipment.
  • the equipment in the protection area can be controlled to slow down or stop running, and the warning equipment can be controlled to output warning information, such as displaying warning words, flashing warning lights, and continuously emitting warning sounds.
  • the image or video of the human body entering the protected area is saved for reference.
  • the application scenarios of the present application can be applied in the industrial field, the security field, and the traffic field to detect the crossing of the human body.
  • the industrial field delineate dangerous areas to prevent people from entering.
  • the equipment can be controlled to pause and warn; for the security field, the entrance and exit of confidential units are demarcated, and when someone enters, they will be warned, photographed and videotaped; or
  • the method of the present application can also be applied to other scenarios that need to detect the entry of a human body.
  • the whole process of the human body cross-border detection method can include the following: first, set the protection boundary according to the protection area, set the inside and outside of the protection boundary to belong to the protection side or the non-protection side respectively, and set the crossing of the human body into the protection boundary condition. Then determine the deployment location and number of cameras according to the protection boundary. After that, the video and images collected by the camera are analyzed to detect the key points of the human body. Then combine the key points of the human body and the out-of-boundary conditions to judge whether the human body is out-of-bounds. After that, control the machines or equipment in the protected area, such as stopping or slowing down the equipment, and setting relevant warning information. Based on this, to solve the problem of inaccurate detection when the human body crosses the boundary, it can automatically detect the human body crossing the boundary, and can accurately detect the human body crossing the boundary, reducing the situation of false detection and missed detection.
  • this application judges the transgression of the human body in the protected area by combining the transgressive conditions and the key points of the human body. Compared with the prior art, it not only realizes automatic monitoring of whether someone intrudes or transgresses in the protected area, but also reduces manual labor. The workload of monitoring can be reduced, and it can accurately detect the cross-border situation of the human body, thereby reducing misjudgments, false positives, and missed negatives.
  • the application can determine the deployment position and number of cameras according to the protection boundary, and at least two cameras deployed can accurately reflect the demarcated boundary and the relative position of the human body next to it in the image, avoiding image damage due to inconsistency between the angle of view and the boundary. cannot accurately reflect the relative position of the boundary line and the human body.
  • step 101 may include but not limited to the following sub-steps:
  • Step 1011 determine the trend setting of the protection boundary according to the protection area
  • Step 1012 determining the protection side and the non-protection side of the protection boundary according to the protection area
  • the protected area includes closed area and non-enclosed area.
  • the protection area includes a closed area and a non-enclosed area.
  • the closed area can be a closed polygon or a closed concave shape, and the protection boundary fits the area that needs to be protected.
  • the protected area; the non-enclosed area is shown in Figure 2C, such as a straight line or a broken line, which can define a certain side of the protected boundary as the protected area, which is conducive to the detection of cross-border at places such as entrances and exits. Therefore, for the protection area, the direction of the protection boundary can be set according to the shape and needs of the actual site.
  • the inside and outside of the protection boundary can be defined according to the protection area as the protection side and the non-protection side, respectively.
  • step 103 may include but not limited to the following sub-steps:
  • Step 1031 determine the endpoint position of the protection boundary according to the protection area
  • Step 1032 determine the deployment position of the camera according to the endpoint position of the protection boundary and the coverage of the camera's field of view.
  • the deployment position of the camera may be determined according to the endpoint position of the protection boundary and the coverage of the camera's field of view. For example, when the protection boundary is determined according to the protection area, cameras can be deployed at each end point of the protection boundary. When the field of view coverage of the camera deployed at the end point cannot completely cover the protection boundary between the end points, the Add cameras between them to ensure that the coverage of the camera's field of view can completely cover the protection boundary.
  • the boundary of the rectangular closed protection area has four endpoints, then four cameras A, B, C, and D can be deployed at the positions of the four endpoints, and camera A , B, C, and D can cover the entire protection boundary; as shown in Figure 9, taking the long linear non-closed protection area as an example, camera A and camera B are respectively deployed at both ends of the boundary, but due to the The field of view coverage of camera A and camera B cannot completely cover the protection boundary between endpoints. Therefore, as shown in Figure 10, it is necessary to add camera C between camera A and camera B so that cameras A, B, and C can capture The coverage of the field can cover the entire protection boundary.
  • the setting method of a single camera looking sideways or squinting will cause false alarms in which people are judged to have crossed the boundary without entering the protection area, and false alarms in which the human body is judged to have not entered the protection area when it actually entered the protection area.
  • cameras are deployed at each endpoint of the protection boundary, so that the images taken by each camera can accurately reflect the relative position of the demarcated boundary and the human body next to it, avoiding the fact that the image cannot accurately reflect the boundaries and boundaries due to inconsistent viewing angles and boundaries.
  • the relative position of the human body overcomes the problems of false positives and false negatives of the human body crossing the boundary existing in the prior art. It should be noted that the number of cameras deployed can be determined by the deployment location of the cameras.
  • step 104 may include but not limited to the following sub-steps:
  • Step 1041 adjusting the coverage height between the fields of view of any two adjacent target cameras to be greater than the height of the human body
  • Step 1042 adjusting the extension of the shooting field of view of the target camera to be greater than a preset threshold
  • the coverage height is the height of the human body in the protection boundary captured by the camera
  • the extension length is the horizontal extension length of the process of the human body entering the protection boundary captured by the camera.
  • the coverage height of the field of view of each camera must meet certain conditions. This is because once the coverage height of the camera is insufficient, the human body cannot be completely captured by the camera on the protection boundary. Therefore, it is necessary to adjust the coverage height between the shooting fields of view of any two adjacent cameras to be greater than the height of the human body. At the same time, in order to ensure that the whole human body can be photographed by the camera when it crosses the boundary from the non-protected side to the protected side, the extension of the field of view of each camera must meet certain conditions.
  • the preset threshold is generally set to one meter. It should be noted that, as shown in Figure 3, as long as the height of the intersection point of the field of view of any adjacent camera is greater than the minimum coverage height, and the minimum coverage height is generally the height of the human body, the distance between the fields of view of any two adjacent cameras is The coverage height H of all meet the requirements of the minimum coverage height. Similarly, as long as the extension length L of any camera is greater than a preset threshold, for example, one meter, the extension lengths of each camera meet the minimum extension length requirement. It should be pointed out that if the minimum coverage height of each camera does not reach the defined minimum coverage height, cameras can be added at or near the coverage intersection of two cameras to increase the coverage height.
  • the extension length is Le.
  • the field of view of the camera itself horizontal field of view Fovh, vertical field of view Fovv.
  • the height of the left and right (horizontal) sides of the camera remains the same, and it is not tilted.
  • the tilt angle can be adjusted front and rear (longitudinal) of the camera.
  • the camera can also be rotated horizontally. Rotate counterclockwise around the x-axis by an angle ⁇ x. Rotate counterclockwise around the y-axis by an angle ⁇ y. Rotate counterclockwise around the z-axis by an angle ⁇ z.
  • the plane perpendicular to the optical axis and with a distance of 1 is Pc.
  • the four edges of the camera’s perspective intersect with this plane, forming four vertices A1, A2, A3, and A4 in the counterclockwise direction of the optical axis.
  • the designer observes the coverage effect and adjusts the parameters to ensure that the extension length requirements are met.
  • the method for calculating the overall coverage height is as follows:
  • intersection height Pch of each adjacent camera above the protection boundary in turn, obtain the intersection height list [Pch1, Pch2...], and compare the height in the list with the configured minimum coverage height Hmin. If the height of a certain Pch cannot reach the defined minimum coverage height, cameras can be added at or near the coverage intersection of two cameras to increase the coverage height.
  • the designer can more accurately determine the position and angle of the camera, which reduces the workload, improves work efficiency, and reduces losses caused by rework and additional materials.
  • the relative position of the demarcated boundary and the nearby human body can be accurately reflected in the image, avoiding that the relative position of the boundary line and the human body cannot be accurately reflected in the image due to inconsistency between the viewing angle and the boundary line.
  • the flexible definition of the human body cross-border standard can be realized. For example, it supports the rules that the foot crossing the boundary line is considered to be out of bounds, and the hand and head crossing the boundary will not call the police.
  • a rectangular enclosed area is protected.
  • the area around equipment in a factory is protected.
  • a Set the protection area parameters. Set the coordinates of the four endpoints, and connect the endpoints to form a protection area. Define the inside and outside of the protection zone.
  • out-of-bounds conditions For example, it is set so that all parts of the human body cannot cross the boundary.
  • control device will suspend the operation, control the warning device to send out warning information, and store the out-of-boundary images and videos.
  • an unenclosed area is protected.
  • a Set area protection parameters. Set the coordinates of the three endpoints to form two connecting lines as boundaries. Define the inside and outside of the protection zone.
  • j According to the definition of the inside and outside of the boundary and the setting of the boundary conditions, extract the coordinates of each key point of the human body, compare the coordinates of the boundary line, and judge whether the key points of the human body except the hand have crossed the boundary line and entered the protection area. If it enters the protected area, it is considered to be out of bounds.
  • control the warning device to send out warning information, and store the out-of-boundary images and videos.
  • the embodiment of the present application also provides a detection device for a human body crossing a boundary.
  • the device for detecting a human body crossing a boundary may include: one or more processors and memory, one processor and memory are taken as an example in FIG. 11 .
  • the processor and the memory may be connected through a bus or in other ways, and connection through a bus is taken as an example in FIG. 11 .
  • the human body crossing detection device is externally connected to at least one camera, and the connection method can be wireless or wired, and the human body crossing detection device can control the operation of the camera.
  • the memory can be used to store non-transitory software programs and non-transitory computer-executable programs, such as the method for detecting human body crossing the boundary in the above-mentioned embodiments of the present application.
  • the processor executes the non-transitory software program and the program stored in the memory, so as to realize the detection method of the human body crossing the boundary in the above-mentioned embodiments of the present application.
  • the memory can include a program storage area and a data storage area, wherein the program storage area can store the operating system and at least one application program required by the function; data etc.
  • the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid-state storage devices.
  • the memory may optionally include a memory that is remotely located relative to the processor, and these remote memories may be connected to the body crossing detection device through a network. Examples of the aforementioned networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
  • the non-transitory software programs and programs required to realize the detection method of human body crossing in the above embodiment of the present application and the program are stored in the memory.
  • the detection of human body crossing in the above embodiment of the present application is executed.
  • the detection method for example, executes method steps 101 to 106 in Fig. 1 described above, method steps 1011 to 1012 in Fig. 5, method steps 1031 to 1032 in Fig. 6, method steps 1041 to 1032 in Fig.
  • Step 1042 determine the protection boundary of the protection area, generate the cross-border condition corresponding to the protection boundary, calculate the deployment position of the camera according to the protection boundary, and adjust the coverage of the shooting field of view of the target camera according to the deployment position of the camera, so that the protection boundary falls within the coverage , read the human body image captured by the target camera, extract the key points of the human body image, and detect the cross-border situation of the human body in the protected area according to the out-of-boundary conditions and key points of the human body.
  • the application can flexibly set the cross-border conditions for the human body to enter the protection boundary, extract the key points of the human body from the human body image captured by the camera, and judge the cross-border situation of the human body in the protected area by combining the cross-border conditions and the key points of the human body.
  • it not only realizes the automatic monitoring of whether someone breaks in or crosses the boundary in the protection area, reduces the workload of manual monitoring, but also can accurately detect the human body crossing the boundary, thereby reducing misjudgment, false positives and missed negatives. occur.
  • the application can determine the deployment position of the camera according to the protection boundary. Accurately deployed multiple cameras can accurately reflect the demarcated boundary and the relative position of the nearby human body in the image, avoiding inaccurate images due to inconsistent viewing angles and boundaries. Reflect the relative position of the boundary line and the human body.
  • the embodiment of the present application also provides an electronic device.
  • the electronic device includes: one or more processors and memories, and one processor and memories are taken as an example in FIG. 12 .
  • the processor and the memory may be connected through a bus or in other ways, and connection through a bus is taken as an example in FIG. 12 .
  • the electronic device is externally connected to at least one camera, and the connection method can be wireless or wired, and the electronic device can control the operation of the camera.
  • the memory can be used to store non-transitory software programs and non-transitory computer-executable programs, such as the method for detecting human body crossing the boundary in the above-mentioned embodiments of the present application.
  • the processor executes the non-transitory software program and the program stored in the memory, so as to realize the detection method of the human body crossing the boundary in the above-mentioned embodiments of the present application.
  • the memory can include a program storage area and a data storage area, wherein the program storage area can store the operating system and at least one application program required by the function; data etc.
  • the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid-state storage devices.
  • the memory may optionally include a memory that is remotely located relative to the processor, and these remote memories may be connected to the body crossing detection device through a network. Examples of the aforementioned networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
  • the non-transitory software programs and programs required to realize the detection method of human body crossing in the above embodiment of the present application and the program are stored in the memory.
  • the detection of human body crossing in the above embodiment of the present application is executed.
  • the detection method for example, executes method steps 101 to 106 in Fig. 1 described above, method steps 1011 to 1012 in Fig. 5, method steps 1031 to 1032 in Fig. 6, method steps 1041 to 1032 in Fig.
  • Step 1042 determine the protection boundary of the protection area, generate the cross-border conditions corresponding to the protection boundary, calculate the deployment position of the camera according to the protection boundary, and adjust the coverage of the shooting field of view of the target camera according to the deployment position of the camera, so that the protection boundary falls into the coverage Range, read the human body image captured by the target camera, extract the key points of the human body image, and detect the cross-border situation of the human body in the protection area according to the cross-border conditions and key points of the human body.
  • the application can flexibly set the cross-border conditions for the human body to enter the protection boundary, extract the key points of the human body from the human body image captured by the camera, and judge the cross-border situation of the human body in the protected area by combining the cross-border conditions and the key points of the human body.
  • it not only realizes the automatic monitoring of whether someone breaks in or crosses the boundary in the protection area, reduces the workload of manual monitoring, but also can accurately detect the human body crossing the boundary, thereby reducing misjudgment, false positives and missed negatives. occur.
  • the application can determine the deployment position of the camera according to the protection boundary. Accurately deployed multiple cameras can accurately reflect the demarcated boundary and the relative position of the nearby human body in the image, avoiding inaccurate images due to inconsistent viewing angles and boundaries. Reflect the relative position of the boundary line and the human body.
  • the embodiment of the present application also provides a computer-readable storage medium, the computer-readable storage medium stores a computer-executable program, and the computer-executable program is executed by one or more control processors, for example, shown in FIG. 11
  • Execution by one of the processors can cause the above-mentioned one or more processors to execute the detection method for human body crossing in the above-mentioned embodiment of the present application, for example, to execute steps 101 to 106 of the method in FIG. 1 described above, and in FIG. 5
  • the application can flexibly set the cross-border conditions for the human body to enter the protection boundary, extract the key points of the human body from the human body image captured by the camera, and judge the cross-border situation of the human body in the protected area by combining the cross-border conditions and the key points of the human body.
  • it not only realizes the automatic monitoring of whether someone breaks in or crosses the boundary in the protection area, reduces the workload of manual monitoring, but also can accurately detect the human body crossing the boundary, thereby reducing misjudgment, false positives and missed negatives. occur.
  • the application can determine the deployment position of the camera according to the protection boundary. Accurately deployed multiple cameras can accurately reflect the demarcated boundary and the relative position of the nearby human body in the image, avoiding inaccurate images due to inconsistent viewing angles and boundaries. Reflect the relative position of the boundary line and the human body.
  • the embodiment of the present application includes: determining the protection boundary of the protection area, generating the cross-border conditions corresponding to the protection boundary, calculating the deployment position of the camera according to the protection boundary, and adjusting the coverage of the field of view of the target camera according to the deployment position of the camera, so that the protection boundary Fall into the coverage area, read the human body image captured by the target camera, extract the key points of the human body image, and detect the cross-border situation of the human body in the protected area according to the cross-border conditions and key points of the human body.
  • the application can flexibly set the cross-border conditions for the human body to enter the protection boundary, extract the key points of the human body from the human body image captured by the camera, and judge the cross-border situation of the human body in the protected area by combining the cross-border conditions and the key points of the human body.
  • it not only realizes the automatic monitoring of whether someone breaks in or crosses the boundary in the protection area, reduces the workload of manual monitoring, but also can accurately detect the human body crossing the boundary, thereby reducing misjudgment, false positives and missed negatives. occur.
  • the application can determine the deployment position of the camera according to the protection boundary. Accurately deployed multiple cameras can accurately reflect the demarcated boundary and the relative position of the nearby human body in the image, avoiding inaccurate images due to inconsistent viewing angles and boundaries. Reflect the relative position of the boundary line and the human body.
  • Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cartridges, tape, magnetic disk storage or other magnetic storage devices, or can Any other medium used to store desired information and which can be accessed by a computer.
  • communication media typically embodies computer readable programs, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media .

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Human Computer Interaction (AREA)
  • Burglar Alarm Systems (AREA)
  • Alarm Systems (AREA)

Abstract

一种人体越界的检测方法、装置和计算机可读存储介质，确定保护区域的保护边界(101)，生成所述保护边界对应的越界条件(102)，根据保护边界计算摄像机的部署位置(103)，根据摄像机的部署位置，调整目标摄像机拍摄视场的覆盖范围，以使得保护边界落入覆盖范围(104)，读取目标摄像机拍摄到的人体图像，提取人体图像的人体关键点(105)，根据越界条件和人体关键点检测人体在保护区域的越界情况(106)。

Description

人体越界的检测方法、装置和计算机可读存储介质
相关申请的交叉引用
本申请基于申请号为202111441251.X、申请日为2021年11月30日的中国专利申请提出，并要求该中国专利申请的优先权，该中国专利申请的全部内容在此引入本申请作为参考。
技术领域
本申请实施例涉及但不限于人体行为检测技术领域，特别是涉及一种人体越界的检测方法、装置和计算机可读存储介质。
背景技术
在工业、交通、安防等各行业中，往往存在着一些危险区域，如机器运行的工位、冶金熔炉等。进入这些区域的人员可能会受到伤害，引发安全事故，同时也可能造成生产的中断。后果非常严重。
目前对于一些有条件安装护栏等保护设施的地方，可通过物理隔离来实现对安全防护的加强。但有的地方不宜安装防护网等设施，主要通过规章制度的宣贯，以及警示牌等方式来提醒人们不要越界。但是人有时会遗忘或者忽略这些信息，导致事故的发生。也有通过摄像机来检测人进入防护区域的方式，但是存在着检测不够准确，未进入保护区域却检测为进入保护区域，或者已进入保护区域，但不能准确检测到的情况，误检误报较多。
发明内容
以下是对本文详细描述的主题的概述。本概述并非是为了限制权利要求的保护范围。
本申请实施例提供了一种人体越界的检测方法、装置和计算机可读存储介质。
第一方面，本申请实施例提供了一种人体越界的检测方法，包括：确定保护区域的保护边界；生成所述保护边界对应的越界条件；根据所述保护边界计算摄像机的部署位置；根据所述摄像机的部署位置，调整目标摄像机拍摄视场的覆盖范围，以使得所述保护边界落入所述覆盖范围；读取所述目标摄像机拍摄到的人体图像，提取所述人体图像的人体关键点；根据所述越界条件和所述人体关键点检测人体在所述保护区域的越界情况。
第二方面，本申请实施例提供了一种人体越界的检测装置，包括：存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述计算机程序时实现如上第一方面所述的人体越界的检测方法。
第三方面，本申请实施例提供了一种电子设备，包括：存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述计算机程序时实现如上第一方面所述的人体越界的检测方法。
第四方面，本申请实施例提供了一种计算机可读存储介质，所述计算机可读存储介质存储有计算机可执行程序，所述计算机可执行程序用于使计算机执行如上第一方面所述的人体越界的检测方法。
本申请的其它特征和优点将在随后的说明书中阐述，并且，部分地从说明书中变得显而易见，或者通过实施本申请而了解。本申请的目的和其他优点可通过在说明书、权利要求书以及附图中所特别指出的结构来实现和获得。
附图说明
附图用来提供对本申请技术方案的进一步理解，并且构成说明书的一部分，与本申请的实施例一起用于解释本申请的技术方案，并不构成对本申请技术方案的限制。
图1是本申请一个实施例提供的一种人体越界的检测方法的主流程图；
图2A是本申请一个实施例提供的保护区域为封闭多边形的示意图；
图2B是本申请一个实施例提供的保护区域为封闭凹陷形的示意图；
图2C是本申请一个实施例提供的保护区域为非封闭区域的示意图；
图3是本申请一个实施例提供的在保护边界上的摄像机覆盖高度和外延长度示意图；
图4是本申请一个实施例提供的人体关键点检测效果图；
图5是本申请一个实施例提供的一种人体越界的检测方法的子流程图；
图6是本申请一个实施例提供的一种人体越界的检测方法的子流程图；
图7是本申请一个实施例提供的矩形封闭保护区域俯视示意图；
图8是本申请一个实施例提供的一种人体越界的检测方法的子流程图；
图9是本申请一个实施例提供的长线形非封闭保护区域俯视示意图；
图10是本申请一个实施例提供的在长线形非封闭保护区域增加摄像机的俯视示意图；
图11是本申请一个实施例提供的人体越界的检测装置结构示意图；
图12是本申请一个实施例提供的电子设备结构示意图。
具体实施方式
为了使本申请的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本申请进行进一步详细说明。应当理解，此处所描述的实施例仅用以解释本申请，并不用于限定本申请。
应了解，在本申请实施例的描述中，多个(或多项)的含义是两个以上，大于、小于、超过等理解为不包括本数，以上、以下、以内等理解为包括本数。如果有描述到“第一”、“第二”等只是用于区分技术特征为目的，而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量或者隐含指明所指示的技术特征的先后关系。
在工业、交通、安防等各行业中，往往存在着一些危险区域，如机器运行的工位、冶金熔炉等。进入这些区域的人员可能会受到伤害，引发安全事故，同时也可能造成生产的中断。后果非常严重。
目前对于一些有条件安装护栏等保护设施的地方，可通过物理隔离来实现对安全防护的加强。但有的地方不宜安装防护网等设施，主要通过规章制度的宣贯，以及警示牌等方式来提醒人们不要越界。但是人有时会遗忘或者忽略这些信息，导致事故的发生。也有通过摄像机来检测人进入防护区域的方式，但是存在着检测不够准确，未进入保护区域却检测为进入保护区域，或者已进入保护区域，但不能准确检测到的情况，误检误报较多。
针对现有技术中人体越界时检测不准确的问题，本申请实施例提供了一种人体越界的检 测方法、装置和计算机可读存储介质,首先根据保护区域设置保护边界，设置保护边界的内侧和外侧分别属于保护侧或者非保护侧，并设定好人体进入保护边界的越界条件。然后根据保护边界确定摄像机的部署位置。之后对摄像机采集的视频、图像进行分析，检测出人体关键点。然后结合人体关键点和越界条件来判断人体是否越界。基于此，解决人体越界时检测不准确的问题，可自动检测人体越界，并且可精确地检测人体越界情况，减少误检、漏检的情况。基于此，本申请可以灵活设定人体进入保护边界的越界条件，在摄像机拍摄到的人体图像中提取人体关键点，通过结合越界条件和人体关键点来判断人体在保护区域的越界情况，相较于现有技术，不但实现自动化监测保护区域内是否有人闯入或越界的情况，减少人工监控的工作量，而且能够精确检测人体越界情况，从而减少误判情况，减少误报、漏报现象的发生。只要人体越过保护界限即可被精确检测到，不会出现摄像机侧视斜视等方式引起的人未进入保护区域却被判断为越界的误报问题，以及人体实际进入保护区域却被判断为未进入的漏报问题。此外，本申请可以根据保护边界确定摄像机的部署位置，准确部署的多台摄像机可在图像中精确地反映划定的界线和旁边人体的相对位置，避免了因视角和界线不一致导致图像中无法准确反映界线和人体的相对位置。调整各个摄像机拍摄视场的覆盖范围，以使得保护边界落入覆盖范围，以此来确定摄像机的覆盖区域，让设计人员更准确的确定摄像机的位置。对于复杂或较大的保护区域，可直观的看到摄像机的覆盖以及各摄像机的交叠情况，可有效的减少纯人工计算覆盖面的工作量，同时可降低人工指定位置的误差，进而削减因摄像机部署不正确导致的返工、追加摄像机、结构件等带来的时间成本、沟通成本、财务成本。
如图1所示，图1是本申请一个实施例提供的一种人体越界的检测方法的流程图。人体越界的检测方法包括但不限于如下步骤：
步骤101，确定保护区域的保护边界；
步骤102，生成保护边界对应的越界条件；
步骤103，根据保护边界计算摄像机的部署位置；
步骤104，根据摄像机的部署位置，调整目标摄像机拍摄视场的覆盖范围，以使得保护边界落入覆盖范围；
步骤105，读取摄像机拍摄到的人体图像，提取人体图像的人体关键点；
步骤106，根据越界条件和人体关键点检测人体在保护区域的越界情况。
可以理解的是，保护区域包括封闭形区域和非封闭形区域，如图2A和图2B所示，封闭形区域可以是封闭的多边形，也可以是封闭的凹陷形状，保护边界则贴合需要保护的保护区域；而非封闭形区域如图2C所示，如一条直线或者折线，可定义保护边界的某一侧为保护区域，有利于出入口等地方的越界检测。因此，对于保护区域可按实际场地的形状以及需要来设置保护边界。
可以理解的是，越界条件可以包括人体部位越界条件和时段越界条件。对于人体部位越界条件，可以针对实际保护场景需要而作出限制特定人体部分的设置，例如，可以将人体部位越界条件设置为手可越界，足不可越界。当然也可以默认设置人体所有部位都不可越界。类似地，对于时段越界条件，同样可以针对实际保护场景需要而作出限制特定时段的设置，例如，可以将时段越界条件设置为白天可越界，晚上不可越界。
可以理解的是，对于摄像机的部署位置，可以保护边界的端点位置和摄像机拍摄视场的覆盖范围来确定。现有技术采用单个摄像机侧视或者斜视的设置方式会引起人未进入保护区 域却被判断为越界的误报问题，以及人体实际进入保护区域却被判断为未进入的漏报问题。而本申请可以将摄像机准确部署在保护边界上适当的位置，使得各个摄像机拍摄的图像中能够精确地反映划定的界线和旁边人体的相对位置，避免了因视角和界线不一致导致图像中无法准确反映界线和人体的相对位置，克服了现有技术中存在的人体越界误报和漏报问题。
可以理解的是，由于每个摄像机都有各自拍摄视场的覆盖范围，可以通过调整摄像机的视场角参数、挂高、朝向角度、光轴方向等来调整摄像机拍摄视场的覆盖范围。将摄像机部署在保护边界的各个端点后，需要调整相应摄像机拍摄视场的覆盖范围，以使得整个保护边界都可以落入覆盖范围内。需要说明的是，在摄像机拍摄的画面上叠加本摄像机覆盖范围内的界线。观察各相邻摄像机的界线是否吻合。如果有偏差，对界线进行调整。消除因摄像机安装的误差导致的相邻摄像机画面内界线不吻合。
可以理解的是，对于调整摄像机拍摄视场的覆盖范围，还包括调整各摄像机的覆盖高度和外延长度。覆盖高度为摄像机拍摄到人体在保护边界的高度，任意相邻两个摄像机拍摄视场之间的覆盖高度一般设置为大于人体高度，以确保界线上的整个人体都可被摄像机覆盖到。外延长度为摄像机拍摄到人体进入保护边界过程的水平延伸长度，此外延长度会延伸到保护边界的非保护侧，摄像机的外延长度一般设置为一米左右，以确保摄像机可以拍摄到整个人体进入保护边界整个过程，即整个人体从保护边界的非保护侧越界走向保护侧的越界过程。需要说明的是，如图3所示，只要任意相邻摄像机视场线相交点的高度都满足大于最低覆盖高度，最低覆盖高度一般为人体高度，则任意相邻两个摄像机拍摄视场之间的覆盖高度H都满足最低覆盖高度的要求。同样地，只要任意摄像机的外延长度L都大于预设阈值，例如一米，则各摄像机的外延长度都满足最小外延长度的要求。
可以理解的是，通过摄像机拍摄人体图像，可以使用深度学习方法从人体图像提取人体关键点。例如，如图4所示，可以使用深度学习算法openpose等检测摄像机拍摄的图像，提取人体关键点。需要指出的是，采用深度学习方法识别人体，可很好的适应光照等环境变化，对光照、地面反光等环境变化有很强的适应能力。
可以理解的是，本申请可以灵活设定人体进入保护边界的越界条件，在摄像机拍摄到的人体图像中提取人体关键点，通过结合越界条件和人体关键点来判断人体在保护区域的越界情况。在一实施方式中，当确定人体关键点满足所述越界条件，判定人体越界进入保护区域内，其中，人体关键点可以包括人的头部、躯干、上肢体、下肢体、手、足等。例如，在工厂内设备周边区域划定一个矩形的保护边界，设置越界条件为禁止晚上人体所有部位进入，因此，当确定任意的人体关键点在晚上进入此矩形的保护边界，则判定人体越界进入工厂设备的保护区域内。
可以理解的是，当判定人体越界进入保护区域内，可以控制保护区域内的设备减缓或者停止运行，并控制警示设备输出警示信息，如显示警示语、警示灯闪烁、持续发出警鸣声等。同时，保存人体进入保护区域内的图像或者视频，以供查阅。
可以理解的是，本申请的应用场景可应用在工业领域、安防领域以及交通领域，对人体越界进行检测。对于工业领域，划定危险区域，防止人员进入，有人进入时可控制设备暂停，并进行警示；对于安防领域，对机密单位的出入口划定界线，在有人进入时进行警示、拍照录像；或者对小区、住宅等的保护，可划定界线，设置夜晚或其它敏感时段，有人进入时警示或保存图像、视频；对于交通领域，对地铁、高铁站台贴近铁轨划定保护区域，有人进入 时进行警示。本申请的方法也可应用到其它需要检测人体进入的场景。
可以理解的是，人体越界的检测方法整个流程可以包括如下：首先根据保护区域设置保护边界，设置保护边界的内侧和外侧分别属于保护侧或者非保护侧，并设定好人体进入保护边界的越界条件。然后根据保护边界确定摄像机的部署位置和部署数量。之后对摄像机采集的视频、图像进行分析，检测出人体关键点。然后结合人体关键点和越界条件，判断人体是否越界。之后，对保护区域内的机器或者设备进行控制，如停止或者减缓设备，以及设置相关警示信息。基于此，解决人体越界时检测不准确的问题，可自动检测人体越界，并且可精确地检测人体越界情况，减少误检、漏检的情况。
可以理解的是，本申请通过结合越界条件和人体关键点来判断人体在保护区域的越界情况，相较于现有技术，不但实现自动化监测保护区域内是否有人闯入或越界的情况，减少人工监控的工作量，而且能够精确检测人体越界情况，从而减少误判情况，减少误报、漏报现象的发生。只要人体越过保护界限即可被精确检测到，不会出现摄像机侧视斜视等方式引起的人未进入保护区域却被判断为越界的误报问题，以及人体实际进入保护区域却被判断为未进入的漏报问题。此外，本申请可以根据保护边界确定摄像机的部署位置和部署数量，部署的至少两台摄像机可在图像中精确地反映划定的界线和旁边人体的相对位置，避免了因视角和界线不一致导致图像中无法准确反映界线和人体的相对位置。调整各个摄像机拍摄视场的覆盖范围，以使得保护边界落入覆盖范围，以此来确定摄像机的覆盖区域，让设计人员更准确的确定摄像机的位置。对于复杂或较大的保护区域，可直观的看到摄像机的覆盖以及各摄像机的交叠情况，可有效的减少纯人工计算覆盖面的工作量，同时可降低人工指定位置的误差，进而削减因摄像机部署不正确导致的返工、追加摄像机、结构件等带来的时间成本、沟通成本、财务成本。
如图5所示，步骤101可以包括但不限于如下子步骤：
步骤1011，根据保护区域确定保护边界的走向设置；
步骤1012，根据保护区域确定保护边界的保护侧和非保护侧；
其中，保护区域包括封闭形区域和非封闭形区域。
可以理解的是，保护区域包括封闭形区域和非封闭形区域，如图2A和图2B所示，封闭形区域可以是封闭的多边形，也可以是封闭的凹陷形状，保护边界则贴合需要保护的保护区域；而非封闭形区域如图2C所示，如一条直线或者折线，可定义保护边界的某一侧为保护区域，有利于出入口等地方的越界检测。因此，对于保护区域可按实际场地的形状以及需要来设置保护边界的走向。同时，可以根据保护区域来定义保护边界的内侧和外侧分别为保护侧和非保护侧。
如图6所示，步骤103可以包括但不限于如下子步骤：
步骤1031，根据保护区域确定保护边界的端点位置；
步骤1032，根据保护边界的端点位置和摄像机拍摄视场的覆盖范围确定摄像机的部署位置。
可以理解的是，对于摄像机的部署位置，可以根据保护边界的端点位置和摄像机拍摄视场的覆盖范围来确定。例如，在根据保护区域确定好保护边界的情况下，可以将摄像机部署在保护边界的各个端点，当部署在端点的摄像机的视场覆盖范围不能完全覆盖端点之间的保护边界，则可以在端点之间增设摄像机，以保证摄像机的视场覆盖范围能够完全覆盖到保护 边界。如图7所示，以矩形封闭保护区域为例，矩形封闭保护区域的边界具有四个端点，则可以将四台摄像机A、B、C、D分别部署在四个端点的位置上，摄像机A、B、C、D拍摄视场的覆盖范围可以覆盖整个保护边界；如图9所示，以长线形非封闭保护区域为例，摄像机A和摄像机B分别部署在边界的两端，但由于摄像机A和摄像机B的视场覆盖范围不能完全覆盖端点之间的保护边界，因此，如图10所示，需要在摄像机A和摄像机B之间增设摄像机C，以使得摄像机A、B、C拍摄视场的覆盖范围可以覆盖整个保护边界。现有技术采用单个摄像机侧视或者斜视的设置方式会引起人未进入保护区域却被判断为越界的误报问题，以及人体实际进入保护区域却被判断为未进入的漏报问题。而本申请将摄像机部署在保护边界的各个端点，使得各个摄像机拍摄的图像中能够精确地反映划定的界线和旁边人体的相对位置，避免了因视角和界线不一致导致图像中无法准确反映界线和人体的相对位置，克服了现有技术中存在的人体越界误报和漏报问题。需要说明的是，根据摄像机的部署数量可以由摄像机的部署位置来确定，例如，如图7所示的矩形保护区域，当确定矩形保护区域的端点为摄像机的部署位置，则可以部署4台摄像机A、B、C、D到四个端点上；如图9所示的长线形非封闭保护区域，除了部署2台摄像机A、B到两个端点上，如图10所示，还需要在端点之间部署1台摄像机C，以覆盖整个保护边界。
如图8所示，步骤104可以包括但不限于如下子步骤：
步骤1041，调整任意相邻两个目标摄像机拍摄视场之间的覆盖高度大于人体高度；
步骤1042，调整目标摄像机拍摄视场的外延长度大于预设阈值；
其中，覆盖高度为摄像机拍摄到人体在保护边界的高度，外延长度为摄像机拍摄到人体进入保护边界过程的水平延伸长度。
可以理解的是，为了保证整个人体在保护边界上都可以被摄像机拍摄到，各个摄像机拍摄视场的覆盖高度必须满足一定条件。这是由于一旦摄像机的覆盖高度不足，会导致人体在保护边界上无法被摄像机完整拍摄到。因此，需要调整任意相邻两个摄像机拍摄视场之间的覆盖高度大于人体高度。同时，为了保证整个人体从保护边界的非保护侧进入到保护侧的越界过程都可以被摄像机拍摄到，各个摄像机拍摄视场的外延长度必须满足一定条件。因此，需要调整各个摄像机拍摄视场的外延长度大于预设阈值，预设阈值一般设置为一米。需要说明的是，如图3所示，只要任意相邻摄像机视场线相交点的高度都满足大于最低覆盖高度，最低覆盖高度一般为人体高度，则任意相邻两个摄像机拍摄视场之间的覆盖高度H都满足最低覆盖高度的要求。同样地，只要任意摄像机的外延长度L都大于预设阈值，例如一米，则各摄像机的外延长度都满足最小外延长度的要求。需要指出的是，如果各个摄像机覆盖的最低高度达不到定义的最低覆盖高度，可在两个摄像机的覆盖交叉点或附近增加摄像机，提升覆盖高度。
可以理解的是，设计人员设置摄像机的视场角、挂高、位置、光轴、外延长度等信息，自动计算摄像机的覆盖范围，供设计人员观察效果。计算摄像机覆盖范围可以采取如下方法：
设定z轴方向为由下向上，x方向由左向右，y轴由近到远。
摄像机坐标为C＝[Cx,Cy,Cz]。则挂高H为Cz。外延长度为Le。摄像机自身的视场角，水平视场角Fovh,垂直视场角Fovv。
一般而言，摄像机的左右侧(横向)高度保持一致，不倾斜。摄像机前后(纵向)可调整倾斜角度。摄像机也可水平旋转。绕x轴逆时针旋转角度为θx。绕y轴逆时针旋转角度为 θy。绕z轴逆时针旋转角度为θz。
根据旋转角度和摄像机坐标，计算旋转矩阵R和平移矩阵T。
在摄像机坐标系内，和光轴方向垂直，距离为1的平面为Pc，摄像机视角的四个边线和此平面相交，按光轴逆时针方向，形成四个顶点A1、A2、A3、A4。四个顶点A1、A2、A3、A4在摄像机坐标系内的坐标依次分别为A1＝[-tan(Fovh),tan(Fovv),1],A2＝[-tan(Fovh),-tan(Fovv),1],A3＝[tan(Fovh),tan(Fovv),1],A4＝[tan(Fovh),tan(Fovv),1]。
计算四个点在世界坐标系内的坐标。Aw＝RA+T。
然后根据四个点在世界坐标系的坐标Aw，以及摄像机坐标C，计算四条视角线L1、L2、L3、L4。然后计算四条线和水平面的交点Ph。
计算交点前，可判断线条是否和水平面水平。如果水平，则无交点。
在一实施方式中，获得四个水平交点Ph后，将四点连接为四条线GL＝[GL1、GL2、GL3、GL4]，形成摄像机在水平面的覆盖区域。然后以摄像机坐标位置z轴为0的点为圆心，Le为半径产生圆形Cy，计算四条线GL和此圆形Cy是否存在交点。如果存在，则不满足最小外延长度。需要继续调整摄像机位置和姿态。
在另一实施方式中，获得四个水平交点Ph后，将四点连接为四条线GL＝[GL1、GL2、GL3、GL4]，形成摄像机在水平面的覆盖区域。然后判断摄像机坐标位置z轴为0的点是否在此覆盖区域内。如果不在，继续调整摄像机使其覆盖此点。
设计人员观察覆盖效果，调整参数确保符合外延长度要求。
继续设置其它各摄像机，计算各摄像机的覆盖范围。
计算整体覆盖的最低高度。计算整体覆盖高度的方法如下：
获取相邻的两个摄像机C1、C2的坐标，以及旋转角度θx、θy、θz，计算四条视场边线L1、L2、L3、L4。将四条边线按相邻关系分为四组，为G1＝[L1,L2]，G2＝[L2,L3]，G3＝[L3,L4]，G4＝[L4,L1]。根据两条线计算所在的平面，为P1、P2、P3、P4。可获得摄像机C1、C2各自的四个视角边面。据此计算C1、C2相邻面的交线，并截取交线和本摄像机的L1、L2、L3、L4相交点范围内的部分。计算此部分交线和C1、C2所在垂直面的交点Pc，获取交点高度Pch。此高度为相邻摄像机交叉覆盖部分在相邻摄像机间保护界线上方的高度。
依次计算各相邻摄像机在保护界线上方的交叉点高度Pch，获得交叉点高度列表[Pch1、Pch2...]，比较列表内高度和配置的最低覆盖高度Hmin。如果某个Pch的高度达不到定义的最低覆盖高度，可在两个摄像机的覆盖交叉点或附近增加摄像机，提升覆盖高度。
基于此，通过自动计算摄像机覆盖区域，能让设计人员更准确的确定摄像机的位置、角度，降低了工作量，提高了工作效率，减少了返工、追加材料等带来的损失。通过将摄像机部署在边界顶点，可在图像中精确的反映划定的界线和旁边人体的相对位置，避免了因视角和界线不一致导致图像中无法准确反映界线和人体的相对位置。通过判断人体的关键点坐标，以及界线、内外部设置等，来判断人体是否越界。相对于人体识别的方式，可更精确的判断人体各部位的越界情况，减少误报。通过人体各部位坐标的检测，可实现人体越界标准的灵活定义。如支持将足部穿过界线认定为越界，手、头越界不报警等越界规则。
以下结合附图和实施例进一步介绍本申请提供的人体越界的检测方法。
如图7所示，对一个矩形封闭的区域进行保护。例如，工厂内设备周边区域。
对应实现越界检测的步骤如下：
a.设置保护区域参数。设置四个端点的坐标，连接端点形成保护区域。定义保护区域内外侧。
b.定义越界条件。如设置为人体各部位都不可越界。
c.计算摄像机覆盖范围。设置摄像机的位置，为各端点。设置最小外延范围、最低覆盖高度。摄像机的视场角参数、挂高、朝向角度，计算摄像机在地面的覆盖范围，结果如图7内A摄像机的虚线范围。
d.计算覆盖范围是否符合外延长度的要求。不符合则进行提示。继续对光轴方向、摄像机参数、位置等进行调整，至符合外延长度要求。
e.继续设置其它各摄像机B、C、D信息，获得各摄像机的覆盖区域。
f.观察保护区域是否已经被全部覆盖。如果已经全部覆盖，则计算各摄像机间的覆盖高度，检查是否符合最低覆盖高度的要求。
最终确定各摄像机位置、光轴方向。
g.按摄像机视场角度、获得的摄像机位置、光轴方向部署摄像机。
观察摄像机拍摄的画面内叠加的界线，对界线进行调整，确保相邻的摄像机画面内界线吻合。
h.使用深度学习算法openpose等检测摄像机拍摄的图像，提取人体关键点。
i.根据界线内外定义，以及越界条件的设置，提取各人体关键点坐标，对比界线坐标，判断是否在界线内部。如果在界线内部，则认为越界。
j.如果判断为越界，则控制设备暂停运行，控制警示设备发出警示信息，存储越界的图像、视频。
如图9所示，对一个非封闭的区域进行保护。例如，高铁站台靠近铁轨的区域。
对应实现越界检测的步骤如下：
a.设置区域保护参数。设置三个端点的坐标，形成两个连接线，作为界线。定义保护区域内外侧。
b.设置越界条件。设置为手可越界，足不可越界。
c.计算摄像机覆盖范围。设置摄像机的位置，为各端点。设置最小外延范围、最低覆盖高度。摄像机的视场角参数、挂高、朝向角度，计算摄像机在地面的覆盖范围，结果如图9内摄像机A和摄像机B的虚线范围。
d.计算覆盖范围是否符合外延长度的要求。不符合则进行提示。继续对光轴方向、摄像机参数、位置等进行调整，至符合外延长度要求。
e.继续设置其它各摄像机信息，获得各摄像机的覆盖区域。
f.观察保护区域是否已经被全部覆盖。如果已经全部覆盖，则计算各摄像机间的覆盖高度，检查是否符合最低覆盖高度的要求。
g.如果覆盖的最低高度达不到定义的最低覆盖高度，可在两个摄像机的覆盖交叉点或附近增加摄像机，提升覆盖高度。如图10所示，在摄像机A和摄像机B之间增加摄像机C。
最终确定各摄像机位置、光轴方向。
h.按摄像机视场角度、获得的摄像机位置、光轴方向部署摄像机。
观察摄像机拍摄的画面内叠加的界线，对界线进行调整，确保相邻的摄像机画面内界线吻合。
i.使用深度学习算法openpose等检测摄像机拍摄的图像，提取人体关键点。
j.根据界线内外定义，以及越界条件的设置，提取各人体关键点坐标，对比界线坐标，判断除手之外的人体关键点是否越过界线，进入保护区域内部。如果进入保护区域，则认为越界。
k.如果判断为越界，则控制警示设备发出警示信息，存储越界的图像、视频。
如图11所示，本申请实施例还提供了一种人体越界的检测装置。
在一实施方式中，该人体越界的检测装置可以包括：一个或多个处理器和存储器，图11中以一个处理器及存储器为例。处理器和存储器可以通过总线或者其他方式连接，图11中以通过总线连接为例。人体越界的检测装置外部与至少一个摄像机连接，连接方式可以是无线或有线，人体越界的检测装置可以控制摄像机运作。
存储器作为一种非暂态计算机可读存储介质，可用于存储非暂态软件程序以及非暂态性计算机可执行程序，如上述本申请实施例中的人体越界的检测方法。处理器通过运行存储在存储器中的非暂态软件程序以及程序，从而实现上述本申请实施例中的人体越界的检测方法。
存储器可以包括存储程序区和存储数据区，其中，存储程序区可存储操作系统、至少一个功能所需要的应用程序；存储数据区可存储执行上述本申请实施例中的人体越界的检测方法所需的数据等。此外，存储器可以包括高速随机存取存储器，还可以包括非暂态存储器，例如至少一个磁盘存储器件、闪存器件、或其他非暂态固态存储器件。在一些实施方式中，存储器可选包括相对于处理器远程设置的存储器，这些远程存储器可以通过网络连接至该人体越界的检测装置。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
实现上述本申请实施例中的人体越界的检测方法所需的非暂态软件程序以及程序存储在存储器中，当被一个或者多个处理器执行时，执行上述本申请实施例中的人体越界的检测方法，例如，执行以上描述的图1中的方法步骤101至步骤106，图5中的方法步骤1011至步骤1012，图6中的方法步骤1031至步骤1032，图8中的方法步骤1041至步骤1042，确定保护区域的保护边界，生成保护边界对应的越界条件，根据保护边界计算摄像机的部署位置，根据摄像机的部署位置调整目标摄像机拍摄视场的覆盖范围，以使得保护边界落入覆盖范围，读取目标摄像机拍摄到的人体图像，提取人体图像的人体关键点，根据越界条件和人体关键点检测人体在保护区域的越界情况。基于此，本申请可以灵活设定人体进入保护边界的越界条件，在摄像机拍摄到的人体图像中提取人体关键点，通过结合越界条件和人体关键点来判断人体在保护区域的越界情况，相较于现有技术，不但实现自动化监测保护区域内是否有人闯入或越界的情况，减少人工监控的工作量，而且能够精确检测人体越界情况，从而减少误判情况，减少误报、漏报现象的发生。只要人体越过保护界限即可被精确检测到，不会出现摄像机侧视斜视等方式引起的人未进入保护区域却被判断为越界的误报问题，以及人体实际进入保护区域却被判断为未进入的漏报问题。此外，本申请可以根据保护边界确定摄像机的部署位置，准确部署的多台摄像机可在图像中精确地反映划定的界线和旁边人体的相对位置，避免了因视角和界线不一致导致图像中无法准确反映界线和人体的相对位置。调整各个摄像机拍摄视场的覆盖范围，以使得保护边界落入覆盖范围，以此来确定摄像机的覆盖区域，让设计人员更准确的确定摄像机的位置。对于复杂或较大的保护区域，可直观的看到摄像机的覆盖以及各摄像机的交叠情况，可有效的减少纯人工计算覆盖面的工作量，同时可降低人工 指定位置的误差，进而削减因摄像机部署不正确导致的返工、追加摄像机、结构件等带来的时间成本、沟通成本、财务成本。
如图12所示，本申请实施例还提供了一种电子设备。
在一实施方式中，该电子设备包括：一个或多个处理器和存储器，图12中以一个处理器及存储器为例。处理器和存储器可以通过总线或者其他方式连接，图12中以通过总线连接为例。电子设备外部与至少一个摄像机连接，连接方式可以是无线或有线，电子设备可以控制摄像机运作。
存储器作为一种非暂态计算机可读存储介质，可用于存储非暂态软件程序以及非暂态性计算机可执行程序，如上述本申请实施例中的人体越界的检测方法。处理器通过运行存储在存储器中的非暂态软件程序以及程序，从而实现上述本申请实施例中的人体越界的检测方法。
存储器可以包括存储程序区和存储数据区，其中，存储程序区可存储操作系统、至少一个功能所需要的应用程序；存储数据区可存储执行上述本申请实施例中的人体越界的检测方法所需的数据等。此外，存储器可以包括高速随机存取存储器，还可以包括非暂态存储器，例如至少一个磁盘存储器件、闪存器件、或其他非暂态固态存储器件。在一些实施方式中，存储器可选包括相对于处理器远程设置的存储器，这些远程存储器可以通过网络连接至该人体越界的检测装置。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
实现上述本申请实施例中的人体越界的检测方法所需的非暂态软件程序以及程序存储在存储器中，当被一个或者多个处理器执行时，执行上述本申请实施例中的人体越界的检测方法，例如，执行以上描述的图1中的方法步骤101至步骤106，图5中的方法步骤1011至步骤1012，图6中的方法步骤1031至步骤1032，图8中的方法步骤1041至步骤1042，确定保护区域的保护边界，生成保护边界对应的越界条件，根据保护边界计算摄像机的部署位置，根据摄像机的部署位置，调整目标摄像机拍摄视场的覆盖范围，以使得保护边界落入覆盖范围，读取目标摄像机拍摄到的人体图像，提取人体图像的人体关键点，根据越界条件和人体关键点检测人体在保护区域的越界情况。基于此，本申请可以灵活设定人体进入保护边界的越界条件，在摄像机拍摄到的人体图像中提取人体关键点，通过结合越界条件和人体关键点来判断人体在保护区域的越界情况，相较于现有技术，不但实现自动化监测保护区域内是否有人闯入或越界的情况，减少人工监控的工作量，而且能够精确检测人体越界情况，从而减少误判情况，减少误报、漏报现象的发生。只要人体越过保护界限即可被精确检测到，不会出现摄像机侧视斜视等方式引起的人未进入保护区域却被判断为越界的误报问题，以及人体实际进入保护区域却被判断为未进入的漏报问题。此外，本申请可以根据保护边界确定摄像机的部署位置，准确部署的多台摄像机可在图像中精确地反映划定的界线和旁边人体的相对位置，避免了因视角和界线不一致导致图像中无法准确反映界线和人体的相对位置。调整各个摄像机拍摄视场的覆盖范围，以使得保护边界落入覆盖范围，以此来确定摄像机的覆盖区域，让设计人员更准确的确定摄像机的位置。对于复杂或较大的保护区域，可直观的看到摄像机的覆盖以及各摄像机的交叠情况，可有效的减少纯人工计算覆盖面的工作量，同时可降低人工指定位置的误差，进而削减因摄像机部署不正确导致的返工、追加摄像机、结构件等带来的时间成本、沟通成本、财务成本。
此外，本申请实施例还提供了一种计算机可读存储介质，该计算机可读存储介质存储有 计算机可执行程序，该计算机可执行程序被一个或多个控制处理器执行，例如，被图11中的一个处理器执行，可使得上述一个或多个处理器执行上述本申请实施例中的人体越界的检测方法，例如，执行以上描述的图1中的方法步骤101至步骤106，图5中的方法步骤1011至步骤1012，图6中的方法步骤1031至步骤1032，图8中的方法步骤1041至步骤1042，确定保护区域保护边界，生成保护边界对应的越界条件，根据保护边界确定摄像机的部署位置，根据摄像机的部署位置，调整目标摄像机拍摄视场的覆盖范围，以使得保护边界落入覆盖范围，读取目标摄像机拍摄到的人体图像，提取人体图像的人体关键点，根据越界条件和人体关键点检测人体在保护区域的越界情况。基于此，本申请可以灵活设定人体进入保护边界的越界条件，在摄像机拍摄到的人体图像中提取人体关键点，通过结合越界条件和人体关键点来判断人体在保护区域的越界情况，相较于现有技术，不但实现自动化监测保护区域内是否有人闯入或越界的情况，减少人工监控的工作量，而且能够精确检测人体越界情况，从而减少误判情况，减少误报、漏报现象的发生。只要人体越过保护界限即可被精确检测到，不会出现摄像机侧视斜视等方式引起的人未进入保护区域却被判断为越界的误报问题，以及人体实际进入保护区域却被判断为未进入的漏报问题。此外，本申请可以根据保护边界确定摄像机的部署位置，准确部署的多台摄像机可在图像中精确地反映划定的界线和旁边人体的相对位置，避免了因视角和界线不一致导致图像中无法准确反映界线和人体的相对位置。调整各个摄像机拍摄视场的覆盖范围，以使得保护边界落入覆盖范围，以此来确定摄像机的覆盖区域，让设计人员更准确的确定摄像机的位置。对于复杂或较大的保护区域，可直观的看到摄像机的覆盖以及各摄像机的交叠情况，可有效的减少纯人工计算覆盖面的工作量，同时可降低人工指定位置的误差，进而削减因摄像机部署不正确导致的返工、追加摄像机、结构件等带来的时间成本、沟通成本、财务成本。
本申请实施例包括：确定保护区域的保护边界，生成保护边界对应的越界条件，根据保护边界计算摄像机的部署位置，根据摄像机的部署位置，调整目标摄像机拍摄视场的覆盖范围，以使得保护边界落入覆盖范围，读取目标摄像机拍摄到的人体图像，提取人体图像的人体关键点，根据越界条件和人体关键点检测人体在保护区域的越界情况。基于此，本申请可以灵活设定人体进入保护边界的越界条件，在摄像机拍摄到的人体图像中提取人体关键点，通过结合越界条件和人体关键点来判断人体在保护区域的越界情况，相较于现有技术，不但实现自动化监测保护区域内是否有人闯入或越界的情况，减少人工监控的工作量，而且能够精确检测人体越界情况，从而减少误判情况，减少误报、漏报现象的发生。只要人体越过保护界限即可被精确检测到，不会出现摄像机侧视斜视等方式引起的人未进入保护区域却被判断为越界的误报问题，以及人体实际进入保护区域却被判断为未进入的漏报问题。此外，本申请可以根据保护边界确定摄像机的部署位置，准确部署的多台摄像机可在图像中精确地反映划定的界线和旁边人体的相对位置，避免了因视角和界线不一致导致图像中无法准确反映界线和人体的相对位置。调整各个摄像机拍摄视场的覆盖范围，以使得保护边界落入覆盖范围，以此来确定摄像机的覆盖区域，让设计人员更准确的确定摄像机的位置。对于复杂或较大的保护区域，可直观的看到摄像机的覆盖以及各摄像机的交叠情况，可有效的减少纯人工计算覆盖面的工作量，同时可降低人工指定位置的误差，进而削减因摄像机部署不正确导致的返工、追加摄像机、结构件等带来的时间成本、沟通成本、财务成本。
本领域普通技术人员可以理解，上文中所公开方法中的全部或某些步骤、系统可以被实 施为软件、固件、硬件及其适当的组合。某些物理组件或所有物理组件可以被实施为由处理器，如中央处理器、数字信号处理器或微处理器执行的软件，或者被实施为硬件，或者被实施为集成电路，如专用集成电路。这样的软件可以分布在计算机可读介质上，计算机可读介质可以包括计算机存储介质(或非暂时性介质)和通信介质(或暂时性介质)。如本领域普通技术人员公知的，术语计算机存储介质包括在用于存储信息(诸如计算机可读程序、数据结构、程序模块或其他数据)的任何方法或技术中实施的易失性和非易失性、可移除和不可移除介质。计算机存储介质包括但不限于RAM、ROM、EEPROM、闪存或其他存储器技术、CD-ROM、数字多功能盘(DVD)或其他光盘存储、磁盒、磁带、磁盘存储或其他磁存储装置、或者可以用于存储期望的信息并且可以被计算机访问的任何其他的介质。此外，本领域普通技术人员公知的是，通信介质通常包含计算机可读程序、数据结构、程序模块或者诸如载波或其他传输机制之类的调制数据信号中的其他数据，并且可包括任何信息递送介质。
以上是对本申请的一些实施进行了说明，但本申请并不局限于上述实施方式，熟悉本领域的技术人员在不违背本申请精神的共享条件下还可作出种种等同的变形或替换，这些等同的变形或替换均包括在本申请权利要求所限定的范围内。

Claims (11)

  1. 一种人体越界的检测方法，包括：
    确定保护区域的保护边界；
    生成所述保护边界对应的越界条件；
    根据所述保护边界计算摄像机的部署位置；
    根据所述摄像机的部署位置，调整目标摄像机拍摄视场的覆盖范围，以使得所述保护边界落入所述覆盖范围；
    读取所述目标摄像机拍摄到的人体图像，提取所述人体图像的人体关键点；
    根据所述越界条件和所述人体关键点检测人体在所述保护区域的越界情况。
  2. 根据权利要求1所述的方法，其中，所述根据所述越界条件和所述人体关键点检测人体在所述保护区域的越界情况，包括：
    当确定所述人体关键点满足所述越界条件，判定人体越界进入所述保护区域内。
  3. 根据权利要求2所述的方法，其中，在所述当确定所述人体关键点满足所述越界条件，判定人体越界进入所述保护区域内之后，还包括：
    控制所述保护区域内的设备减缓或者停止运行，并输出警示信息；和/或，
    保存人体进入所述保护区域内的图像或者视频。
  4. 根据权利要求1所述的方法，其中，所述确定保护区域的保护边界，包括：
    根据保护区域确定所述保护边界的走向设置；
    根据保护区域确定所述保护边界的保护侧和非保护侧；
    其中，所述保护区域包括封闭形区域和非封闭形区域。
  5. 根据权利要求1所述的方法，其中，所述根据所述保护边界计算摄像机的部署位置，包括：
    根据所述保护区域确定所述保护边界的端点位置；
    根据所述保护边界的端点位置和所述摄像机拍摄视场的覆盖范围确定所述摄像机的部署位置。
  6. 根据权利要求5所述的方法，其中，在所述根据所述保护边界的端点位置和所述摄像机拍摄视场的覆盖范围确定所述摄像机的部署位置之后，还包括：
    根据所述摄像机的部署位置确定所述摄像机的部署数量。
  7. 根据权利要求1所述的方法，其中，所述调整目标摄像机拍摄视场的覆盖范围，包括：
    调整任意相邻两个所述目标摄像机拍摄视场之间的覆盖高度大于人体高度；
    调整所述目标摄像机拍摄视场的外延长度大于预设阈值；
    其中，所述覆盖高度为所述摄像机拍摄到人体在所述保护边界的高度，所述外延长度为所述摄像机拍摄到人体进入所述保护边界过程的水平延伸长度。
  8. 根据权利要求1至7任意一项所述的方法，其中，所述越界条件包括人体部位越界条件和时段越界条件。
  9. 一种人体越界的检测装置，包括：存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述计算机程序时实现如权利要求1至8中任意一项所述的人体越界的检测方法。
  10. 一种电子设备，包括：存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述计算机程序时实现如权利要求1至8中任意一项所述的人体越界的检测方法。
  11. 一种计算机可读存储介质，所述计算机可读存储介质存储有计算机可执行程序，所述计算机可执行程序用于使计算机执行如权利要求1至8任意一项所述的人体越界的检测方法。
PCT/CN2022/130621 2021-11-30 2022-11-08 人体越界的检测方法、装置和计算机可读存储介质 WO2023098419A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111441251.XA CN116206250A (zh) 2021-11-30 2021-11-30 人体越界的检测方法、装置和计算机可读存储介质
CN202111441251.X 2021-11-30

Publications (1)

Publication Number Publication Date
WO2023098419A1 true WO2023098419A1 (zh) 2023-06-08

Family

ID=86513413

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/130621 WO2023098419A1 (zh) 2021-11-30 2022-11-08 人体越界的检测方法、装置和计算机可读存储介质

Country Status (2)

Country Link
CN (1) CN116206250A (zh)
WO (1) WO2023098419A1 (zh)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160142640A1 (en) * 2014-11-18 2016-05-19 Canon Kabushiki Kaisha Imaging apparatus, control method of imaging apparatus, and storage medium
CN106327525A (zh) * 2016-09-12 2017-01-11 安徽工业大学 一种机房重地越界行为实时监测方法
CN112818768A (zh) * 2021-01-19 2021-05-18 南京邮电大学 一种基于元学习的变电站改扩建违章行为智能化识别方法
CN113139521A (zh) * 2021-05-17 2021-07-20 中国大唐集团科学技术研究院有限公司中南电力试验研究院 一种用于电力监控的行人越界标监测方法
CN113449675A (zh) * 2021-07-12 2021-09-28 西安科技大学 一种煤矿人员越界检测方法
CN113657309A (zh) * 2021-08-20 2021-11-16 山东鲁软数字科技有限公司 一种基于Adocf的穿越安全围栏违章行为检测方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160142640A1 (en) * 2014-11-18 2016-05-19 Canon Kabushiki Kaisha Imaging apparatus, control method of imaging apparatus, and storage medium
CN106327525A (zh) * 2016-09-12 2017-01-11 安徽工业大学 一种机房重地越界行为实时监测方法
CN112818768A (zh) * 2021-01-19 2021-05-18 南京邮电大学 一种基于元学习的变电站改扩建违章行为智能化识别方法
CN113139521A (zh) * 2021-05-17 2021-07-20 中国大唐集团科学技术研究院有限公司中南电力试验研究院 一种用于电力监控的行人越界标监测方法
CN113449675A (zh) * 2021-07-12 2021-09-28 西安科技大学 一种煤矿人员越界检测方法
CN113657309A (zh) * 2021-08-20 2021-11-16 山东鲁软数字科技有限公司 一种基于Adocf的穿越安全围栏违章行为检测方法

Also Published As

Publication number Publication date
CN116206250A (zh) 2023-06-02

Similar Documents

Publication Publication Date Title
JPH04148298A (ja) 違法駐車検出警告装置
CN105141912B (zh) 一种信号灯重定位的方法及设备
CN102708655A (zh) 一种消防通道占用监测报警方法及其系统
CN112052821B (zh) 消防通道安全检测方法、装置、设备及存储介质
TWI719766B (zh) 警戒區域設定系統及其方法
KR20200132137A (ko) 움직임 감지와 레이더 센서를 이용한 탐지 객체 위치 산출 장치
WO2023098419A1 (zh) 人体越界的检测方法、装置和计算机可读存储介质
KR20220010858A (ko) 차량 우회전시 보행자 알림 시스템
CN104103077A (zh) 一种人头检测方法和装置
BR102015005652A2 (pt) aparelho para controlar operações de formação de imagem de uma câmera óptica, e, sistema para reconhecer um número de veículo de uma placa de licença de um veículo
JP2014026372A (ja) カメラの撮影視野変動検知装置
CN113682319A (zh) 摄像头调整方法及装置、电子设备和存储介质
KR101903615B1 (ko) 영상감시 시스템 및 이를 이용한 영상감시 방법
JP2005070850A (ja) ゲート制御装置及びゲート制御システム
JP2019147443A (ja) 踏切監視装置、画像監視装置、踏切監視方法及び踏切監視プログラム
CN202650193U (zh) 一种消防通道占用监测报警系统
JP6562722B2 (ja) 侵入監視装置及び侵入監視システム
KR102640060B1 (ko) 선박용 모니터링 장치
CN104301676B (zh) 一种监控目标搜索方法和装置以及球型摄像机
KR102434154B1 (ko) 영상감시시스템에서의 고속 이동물체의 위치 및 모션 캡쳐 방법
KR20230020184A (ko) 고정 카메라 및 이동 카메라를 이용한 영상 분석 장치
CN113627343A (zh) 一种行人身高检测方法、装置、设备及可读存储介质
CN112706159B (zh) 机器人控制方法、装置及机器人
TW201447825A (zh) 保全影像辨識系統
KR101520596B1 (ko) 폐쇄회로 텔레비전

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22900224

Country of ref document: EP

Kind code of ref document: A1