Person positioning method, apparatus, device, and storage medium

Abstract

The application provides a personnel positioning method, device, equipment and storage medium, the method comprises the following steps: acquiring image data collected by a front-end collection device arranged in a target area; for any personnel, determining the coordinates of the projection point of the detection box of the personnel on a specified straight line according to the coordinate information of the detection box of the personnel in the monitoring image of the current front-end collection device; and determining the position of the personnel in the target area according to the coordinates of the projection point of the detection box on the specified straight line and the coordinates of the calibration point. The method can improve the intuitiveness of personnel positioning.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a method, apparatus, device and storage medium for personnel positioning. Background Technology

[0002] With the rapid development of video surveillance technology, the application of front-end acquisition devices such as cameras has become increasingly widespread. Based on the image data from these front-end acquisition devices, personnel detection within the monitored area can be achieved.

[0003] How to intuitively locate people within a monitored area has become an urgent technical problem to be solved. Summary of the Invention

[0004] In view of this, this application provides a method, apparatus, device and storage medium for personnel positioning.

[0005] According to a first aspect of the embodiments of this application, a method for locating personnel is provided, comprising:

[0006] Acquire image data collected by front-end acquisition devices deployed in the target area;

[0007] For any person, based on the coordinate information of the detection box in the monitoring image of the current front-end acquisition device, determine the coordinates of the projection point of the detection box on the specified straight line;

[0008] Based on the coordinates of the projection point of the detection frame on the specified straight line and the coordinates of the calibration point, the position of the person in the target area is determined; wherein, the calibration point is the projection point of the preset calibration position on the specified straight line.

[0009] According to a second aspect of the embodiments of this application, a personnel positioning device is provided, comprising:

[0010] The acquisition unit is configured to acquire image data collected by front-end acquisition devices deployed in the target area;

[0011] The determining unit is configured to, for any person, determine the coordinates of the projection point of the detection box on a specified straight line based on the coordinate information of the detection box in the monitoring image of the current front-end acquisition device;

[0012] The positioning unit is configured to determine the position of the person in the target area based on the coordinates of the projection point of the detection frame on a specified straight line and the coordinates of the calibration point; wherein the calibration point is the projection point of a preset calibration position on the specified straight line.

[0013] According to a third aspect of the embodiments of this application, an electronic device is provided, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0014] Memory, used to store computer programs;

[0015] The processor, when executing a program stored in memory, implements the personnel positioning method of the first aspect.

[0016] According to a fourth aspect of the embodiments of this application, a non-transitory computer-readable storage medium is provided, wherein a computer program is stored in the non-transitory computer-readable storage medium, and when the computer program is executed by a processor, it implements the personnel positioning method of the first aspect.

[0017] The personnel positioning method of this application embodiment pre-sets a calibration position for personnel positioning and a straight line (i.e., a designated straight line) for assisting personnel positioning, and determines the coordinates of the projection point (i.e., the calibration point) of the preset calibration position on the designated straight line. For any person, the coordinates of the detection frame in the monitoring image of the current front-end acquisition device can be used to determine the coordinates of the projection point of the detection frame on the designated straight line. Then, the position of the person in the target area can be determined based on the coordinates of the projection point of the detection frame on the designated straight line and the coordinates of the calibration point. Personnel positioning is achieved by calibrating the position, which improves the intuitiveness of personnel positioning, so that relevant personnel can more intuitively determine the position of the person to be located in the target area. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating a monitoring method provided in an embodiment of this application;

[0019] Figure 2 This is a flowchart illustrating a personnel positioning method provided in an embodiment of this application;

[0020] Figure 3A This is a schematic diagram illustrating the coordinates of a streetlight pole appearing in a monitoring screen, as provided in an embodiment of this application.

[0021] Figure 3B This is a schematic diagram of a monitoring screen with marked virtual lines and calibration points provided in an embodiment of this application;

[0022] Figure 4 This is a schematic diagram of the overall process of personnel behavior early warning provided in an embodiment of this application;

[0023] Figure 5 This is a schematic diagram of an algorithm for personnel monitoring provided in an embodiment of this application;

[0024] Figure 6 This is a schematic diagram of the structure of a monitoring device provided in an embodiment of this application;

[0025] Figure 7This is a schematic diagram of the structure of a personnel positioning device provided in an embodiment of this application;

[0026] Figure 8 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0027] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0028] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0029] To enable those skilled in the art to better understand the technical solutions provided in the embodiments of this application, and to make the above-mentioned objectives, features and advantages of the embodiments of this application more apparent and understandable, the technical solutions in the embodiments of this application will be further described in detail below with reference to the accompanying drawings.

[0030] It should be noted that the sequence number of each step in the embodiments of this application does not imply the order of execution. 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 this application.

[0031] Please see Figure 1 This is a flowchart illustrating a monitoring method provided in an embodiment of this application, as shown below. Figure 1 As shown, the monitoring method may include the following steps:

[0032] Step S100: Obtain image data collected by the front-end acquisition devices deployed in the target area; wherein, multiple front-end acquisition devices are deployed in the target area; the acquisition area of ​​each front-end acquisition device is the monitoring area.

[0033] In this application embodiment, the target area does not specifically refer to a fixed area, but can refer to any area that uses the scheme provided in this application embodiment for behavior monitoring.

[0034] Multiple front-end data acquisition devices are deployed in the target area.

[0035] For example, the target area may include pedestrian walkways on both sides of a bridge. For any one side of the pedestrian walkway, multiple front-end data acquisition devices, such as cameras, can be deployed along the direction of traffic.

[0036] In this embodiment of the application, in order to monitor personnel within the target area, image data from front-end acquisition devices deployed in the target area can be acquired.

[0037] For example, the image data may include video data or picture data.

[0038] For example, when the image data is picture data, image data collected by the front-end device deployed in the target area can be acquired at preset time intervals.

[0039] Step S110: For any first person, query the image data of the first person in the current monitoring area; the current monitoring area is the acquisition area of ​​the current front-end acquisition device that captured the first person; the first person is any person identified after performing personnel clustering on the image data.

[0040] In this embodiment of the application, personnel detection and tracking can be performed based on image data collected by each front-end acquisition device in the target area, and the same person under different front-end acquisition devices can be identified by performing personnel clustering on the image data of different front-end acquisition devices.

[0041] For any person identified by personnel clustering of image data (which can be called the first person), the captured data of the first person in the current monitoring area (i.e., the acquisition area of ​​the current front-end acquisition device that captured the first person) can be queried from the acquired image data.

[0042] Step S120: Based on the retrieved snapshot data, determine whether the first person meets the conditions for staying in the monitored area. If yes, proceed to step S130; otherwise, proceed to step S140.

[0043] Step S130: Issue the first warning.

[0044] Step S140: Determine whether to issue an early warning based on the subsequent capture data of the first person.

[0045] In this embodiment of the application, in order to promptly detect the situation of personnel lingering in the monitoring area, if the capture data of the first person in the current monitoring area is obtained in accordance with the method described in the above embodiment, it can be determined whether the first person meets the conditions for lingering in the monitoring area based on the retrieved capture data.

[0046] If the first person meets the conditions for staying in the monitored area, a first warning (also known as a suspected monitoring area stay warning) can be issued to indicate that the first person is staying in the monitored area.

[0047] If the first person does not meet the conditions for staying in the monitored area, it can be determined whether to issue an early warning based on the subsequent capture data of the first person. For example, if the subsequent capture data of the first person determines that the first person meets the conditions for staying in the monitored area, an early warning can be issued. Alternatively, if the subsequent capture data of the first person determines that the first person meets the trigger conditions for other types of early warnings, other types of early warnings can be issued.

[0048] It can be seen that, in Figure 1 In the illustrated method, image data collected by front-end acquisition devices deployed in the target area is acquired, and personnel clustering is performed based on the image data of multiple front-end acquisition devices deployed in the target area to identify the same person (i.e., the first person) under different front-end acquisition devices. For any first person, the captured data of the first person in the current monitoring area is queried from the acquired image data, and it is determined whether the first person meets the conditions for staying in the monitoring area based on the retrieved captured data. If the first person meets the conditions for staying in the monitoring area, a first warning is issued; if the first person does not meet the conditions for staying in the monitoring area, it is determined whether to issue a warning based on the subsequent captured data of the first person. This enables personnel monitoring in scenarios with multiple cameras and cross-camera movement.

[0049] In some embodiments, the conditions for staying in the monitored area may include:

[0050] The first person is captured by the same front-end acquisition device more than a first preset number of times, and the first person is continuously captured by the front-end acquisition device for a period of time that reaches a first preset time.

[0051] For example, in order to promptly detect situations where personnel linger in a certain monitoring area for an extended period of time, the duration of personnel's stay in the same monitoring area can be used as one of the criteria for staying in the monitoring area.

[0052] For example, the duration of a person's stay in the monitoring area (i.e., the acquisition area of ​​the front-end acquisition device) can be characterized by the time a person is continuously captured by the same front-end acquisition device.

[0053] In addition, to avoid false detections leading to false alarms, the monitoring zone stay conditions can also include the number of times a person is detected by the front-end data collection equipment.

[0054] Accordingly, the conditions for staying in the monitoring area may include the number of times a first person is captured by the same front-end acquisition device exceeds a first preset number, and the time that the first person is continuously captured by the front-end acquisition device reaches a first preset time.

[0055] In some embodiments, if the first person meets the conditions for staying in the monitored area, the monitoring method provided in this application embodiment may further include:

[0056] Based on the coordinate information of the detection box in the monitoring image of the current front-end acquisition device, determine the coordinates of the projection point of the detection box on the specified straight line;

[0057] Based on the coordinates of the projection point of the detection frame on the specified straight line and the coordinates of the calibration point, the position of the person in the target area is determined; wherein, the calibration point is the projection point of the preset calibration position on the specified straight line.

[0058] For example, if it is determined that the first person meets the conditions for staying in the monitored area, the location of the first person in the target area can be further determined. The specific implementation process can be found in [reference needed]. Figure 2 The relevant descriptions of the personnel positioning process shown are not repeated here in the embodiments of this application.

[0059] In some embodiments, determining whether to issue an alert based on subsequent snapshots taken by the first person may include:

[0060] If the data of the first person leaving the target area is found, then it is determined that the first person has left the target area.

[0061] For example, it can be determined whether the object to be monitored has left the target area based on image data collected by front-end acquisition devices deployed at the entrances and exits of the target area.

[0062] For example, the front-end acquisition device corresponding to the last capture point of the object to be monitored can be deployed at the entrance and exit of the target area, and the object to be monitored can be determined to be moving away from the target area based on the distribution of the capture points of the object to be monitored.

[0063] In some embodiments, determining whether to issue an alert based on subsequent snapshots taken by the first person may include:

[0064] If no image data of the first person is received from the next front-end acquisition device during the second time period, a suspected blind spot lingering warning will be issued for the first person. The suspected blind spot lingering warning is used to indicate that the first person has entered the monitoring blind spot between the current front-end acquisition device and the next front-end acquisition device.

[0065] For example, if no data is captured by the next front-end acquisition device for any first person during the second continuous time period, it can be determined that the first person is suspected of entering the monitoring blind zone between the current front-end acquisition device and the next front-end acquisition device, and a warning for suspected blind zone stay is issued for the first person to notify relevant personnel that the first person has entered the monitoring blind zone between the current front-end acquisition device and the next front-end acquisition device.

[0066] For example, the next front-end acquisition device refers to the adjacent front-end acquisition device following the current front-end acquisition device according to the movement direction of the first person.

[0067] For example, the direction of movement of the first person can be determined based on the distribution of the first person's capture points in the target area.

[0068] The distribution of capture points for the first person in the target area can be determined based on the image data from each front-end acquisition device.

[0069] For example, suppose cameras 1, 2, 3, ... are deployed sequentially along the direction of the first person's movement. Based on the distribution of the first person's capture points, it can be determined that the first person last appeared in the monitoring range of camera 3. Then, the monitoring blind spot between camera 3 and camera 4 can be identified.

[0070] It should be noted that, in this embodiment of the application, in order to avoid false alarms of suspected blind spot loitering due to false detection, when determining whether the monitored object meets the blind spot loitering conditions for any first person, the number of times the first person is captured can also be considered. That is, for any first person, if it is determined that the number of times the first person is captured exceeds a preset number (which can be called the second preset number), and if no capture data of the first person is received from the next front-end acquisition device for a continuous second time period, a suspected blind spot loitering warning can be issued for the first person.

[0071] In one example, the above-mentioned determination of whether to issue an alert based on subsequent snapshots taken by the first person may further include:

[0072] The system uses a PTZ camera associated with the location of the first person in the blind spot to capture images of that blind spot and determines whether the first person is in that blind spot based on the images.

[0073] For example, in order to more accurately determine whether a person is in a surveillance blind spot, after determining the surveillance blind spot (which can be called the target surveillance blind spot) where the first person is located in the manner described above, the PTZ camera associated with the target surveillance blind spot can be linked to collect images of the target surveillance blind spot, and the first person can be determined to be in the target surveillance blind spot based on the images.

[0074] For example, PTZ cameras can be deployed at higher levels in the target area based on the distribution of blind spots in the front-end acquisition devices. Different preset positions of the PTZ cameras can correspond to different blind spots (one or more PTZ cameras can be deployed in the target area, the specific number depending on the number and distribution of blind spots). Once the first person is located in the target blind spot, the PTZ cameras associated with the target blind spot can be linked to control them to acquire images from the corresponding preset positions to verify whether the first person is in the target blind spot.

[0075] For example, if it is determined from the image captured by the PTZ camera that the first person is in the blind spot of the target monitoring, specific voice information can be played through the voice device, such as a voice message to persuade the first person to leave.

[0076] In some embodiments, determining the distribution of capture points for the first person in the target area based on image data from each front-end acquisition device may include:

[0077] For any front-end acquisition device, based on the image data acquired by the front-end acquisition device, a tracking algorithm is used to determine the distribution of capture points of the first person under the front-end acquisition device;

[0078] For different front-end acquisition devices, structured feature clustering is used to identify the same first person under different front-end acquisition devices;

[0079] Based on the distribution of capture points of the same first person under different front-end acquisition devices, the distribution of capture points of the first person in the target area is determined.

[0080] For example, under the same front-end acquisition device, a tracking algorithm can be used to track people detected based on the image data acquired by the front-end acquisition device, and determine the distribution of capture points of the same monitored object under the front-end acquisition device.

[0081] For different front-end acquisition devices, structured feature clustering can be used to identify the same person under different front-end acquisition devices.

[0082] For example, for front-end acquisition device 1, the detected object can be performed based on the image acquired by front-end acquisition device 1. For a detected person (e.g., person A), the structured features of person A can be extracted and stored. For a person detected based on the image acquired by front-end acquisition device 2, the structured features of that person can be extracted, and the structured features of that person and the structured features of person A can be clustered; if the clustering is successful, it can be determined that the person and person A are the same person.

[0083] Furthermore, based on the distribution of capture points for the same person under different front-end acquisition devices, the distribution of capture points for that person in the target area can be determined.

[0084] In one example, the structured features of a person include the structured features of different parts of the person;

[0085] The above-mentioned structured feature clustering method is used to identify the same first person under different front-end acquisition devices, including:

[0086] For different front-end acquisition devices, a structured feature clustering method is used to cluster the structured features of the same parts of the first person under different front-end acquisition devices, and to identify the same parts of the same first person under different front-end acquisition devices.

[0087] Based on the correlation between different parts of the same first person, the different parts of the same first person are correlated to identify the same first person under different front-end acquisition devices.

[0088] For example, considering the situation where a person moves within the target area, the front-end acquisition device that acquires the image of the person cannot guarantee that it can extract the structural features of all parts of the person based on the image acquired by the front-end acquisition device.

[0089] For example, in some cases, when a person is within the monitoring range of a front-end acquisition device but is facing away from the device, the person's facial structure features may not be extracted from the image captured by the device.

[0090] Accordingly, in order to improve the accuracy of identifying the same person under different front-end acquisition devices, structured features can be extracted from different parts of the person, and the same parts of the same person under different front-end acquisition devices can be identified by clustering the structured features of the same parts. Furthermore, based on the correlation between different parts of the same person, the same person under different front-end acquisition devices can be identified.

[0091] For example, different body parts can include the face and the body.

[0092] For example, for different front-end acquisition devices, a structured feature clustering method is used to cluster the structured features of the same parts of the personnel under different front-end acquisition devices, and to identify the same parts of the same personnel under different front-end acquisition devices.

[0093] For example, clustering algorithms may include, but are not limited to, K-means clustering, hierarchical clustering, and DBSCAN (Density-Based Spatial Clustering of Applications with Noise).

[0094] As an example, the above-mentioned method of using structured feature clustering to cluster the structured features of the same body parts of the first person under different front-end acquisition devices, and to identify the same body parts of the same first person under different front-end acquisition devices, may include:

[0095] For any face structured feature under any front-end acquisition device, the face structured feature is clustered with the stored baseline face structured features to determine the target baseline face structured feature that is successfully clustered with the face structured feature, and a unique personnel ID is configured for the face structured feature based on the target baseline face structured feature.

[0096] as well as,

[0097] For any human structured feature acquired by a front-end acquisition device, the human structured feature is clustered with the stored baseline human structured features to determine the target baseline human structured feature that is successfully clustered with the human structured feature.

[0098] For example, for any front-end acquisition device, personnel detection can be performed based on the image data acquired by the front-end acquisition device.

[0099] When a person is detected, their facial structure features and / or body structure features can be extracted separately.

[0100] For example, facial structural features may include, but are not limited to, facial contours, eye position, eyebrow shape, lip shape, and nose features.

[0101] Human structural features may include, but are not limited to, height, body shape, head position, and limb posture.

[0102] For any face structured feature under any front-end acquisition device, the face structured feature can be clustered with the stored baseline face structured features to determine the baseline face structured feature that is successfully clustered with the face structured feature (referred to as the target baseline face structured feature in this paper), and a unique personnel ID can be configured for the face structured feature based on the target baseline face structured feature.

[0103] It should be noted that, in the embodiments of this application, if there is no stored baseline face structured feature, or if there is no target baseline face structured feature that is successfully clustered with the face structured feature, the face structured feature can be used as the baseline face structured feature and stored.

[0104] Furthermore, the stored baseline facial structured features can be updated using facial structured features of the same person that are of higher quality (e.g., those with higher scores determined by a pre-defined scoring algorithm).

[0105] Furthermore, when adding new stored baseline facial structured features, a unique personnel ID can be assigned to the baseline facial structured feature, and updating the baseline facial structured feature does not change the unique personnel ID; in subsequent processes, for facial structured features that are successfully clustered with the baseline facial structured feature, the configured unique personnel ID can be the unique facial ID of the baseline facial structured feature.

[0106] Similarly, for any face structured feature from any front-end acquisition device, the face structured feature can be clustered with the stored baseline face structured features to determine the target baseline human structured feature that is successfully clustered with the face structured feature.

[0107] It should be noted that, in the embodiments of this application, if there is no stored baseline human structured feature, or if there is no target baseline human structured feature that is successfully clustered with the human structured feature, the human structured feature can be used as the baseline human structured feature and stored.

[0108] Furthermore, considering that the structural features of the human body may vary significantly depending on its orientation, when storing the baseline structural features, the baseline structural features for different orientations can be stored separately to increase the probability of successful clustering.

[0109] For example, if the identification of the same part of the same person under different front-end acquisition devices is achieved in the above manner, the different parts of the same person can be associated based on the correlation between different parts of the same person to determine the same person under different front-end acquisition devices.

[0110] As an example, the above-mentioned method of associating different body parts of the same person based on the correlation between them, and identifying the same person under different front-end data acquisition devices, can include:

[0111] Based on the face-body association ID corresponding to the target baseline face structured features and the face-body association ID corresponding to the target baseline body structured features, the face structured features and body structured features are aggregated to identify the same person under different front-end acquisition devices; the face-body association ID is used to associate the person and body of the same person.

[0112] For example, for any front-end acquisition device and any detected person corresponding to the front-end acquisition device, if no target baseline face structured features and / or target body structured features that have been successfully clustered are found based on the person's face structured features and / or body structured features, the baseline face structured features and baseline body structured features of the person can be obtained based on the face structured features and body structured features extracted from the image data of the front-end acquisition device.

[0113] For example, a tracking algorithm can be used to determine the distribution of detection points of the person under the front-end acquisition device, and the person's facial structured features and human body structured features can be extracted based on the image data of the person under the front-end acquisition device. The extracted facial structured features and human body structured features are stored as reference facial structured features and reference human body structured features, and the same facial and human body association ID is assigned to the reference facial structured features and reference human body structured features.

[0114] In subsequent processes, for other facial structured features that are successfully clustered with the benchmark facial structured features, and for other human structured features that are successfully clustered with the benchmark human structured features, the facial-human association ID can be configured. This enables the association between facial structured features and human structured features, allowing the identification of the facial structured features and human structured features of the same person. Consequently, it enables the identification of the same person under different front-end acquisition devices.

[0115] It should be noted that, in the embodiments of this application, since the same front-end acquisition device can be used to determine the distribution of detection points of the same person under the same front-end acquisition device, that is, the detection of the same person in different frames of image data of the front-end acquisition device can be determined by the tracking algorithm. Therefore, for any person, if the baseline structural features of the person under the front-end acquisition device are determined by the method of structural feature clustering, it is not necessary to perform structural feature clustering again based on the structural features of the person under the front-end acquisition device (in the case of clustering by part, the structural features of different parts need to be successfully clustered once).

[0116] Please see Figure 2 This is a flowchart illustrating a personnel positioning method provided in an embodiment of this application, as shown below. Figure 2 As shown, the personnel location method may include the following steps:

[0117] Step S200: Obtain image data collected by the front-end acquisition device deployed in the target area.

[0118] Step S210: For any person, based on the coordinate information of the detection box in the monitoring image of the current front-end acquisition device, determine the coordinates of the projection point of the detection box on the specified straight line.

[0119] Step S220: Determine the position of the person in the target area based on the coordinates of the projection point of the detection frame on the specified straight line and the coordinates of the calibration point; wherein, the calibration point is the projection point of the preset calibration position on the specified straight line.

[0120] In this embodiment of the application, in order to determine the position of a person in the target area, multiple calibration positions for positioning the person can be preset, and the coordinates of the projection point (which can be called the calibration point) of the preset calibration position on the specified straight line can be determined.

[0121] For example, the designated straight line is a preset straight line used to assist personnel in positioning.

[0122] It should be noted that, unless otherwise specified, the coordinate values ​​mentioned in the embodiments of this application can be pixel coordinate values ​​in the screen coordinate system, such as coordinate values ​​in a coordinate system with the top left corner of the screen as the origin, the horizontal direction to the right as the positive x-axis, and the vertical direction downward as the positive y-axis.

[0123] For any person, such as Figure 1 The first person in the method flow shown can determine the coordinates of the projection point of the detection box on the specified straight line based on the coordinate information of the detection box in the monitoring image of the current front-end acquisition device.

[0124] For example, the coordinates of the projection point of the detection box on a specified straight line can be determined based on the coordinate information of the latest detection box in the monitoring image of the current front-end acquisition device.

[0125] For example, the current front-end acquisition device can be the front-end acquisition device that currently captures the person's image.

[0126] For example, for any detection box, the coordinates of the projection point of the detection box on the specified line can be the coordinates of the projection point of a certain edge of the detection box on the specified line, or the coordinates of the projection point of a certain vertex of the detection box on the specified line, or the coordinates of the projection point of the center point of the detection box on the specified line.

[0127] In this embodiment of the application, the position of the person in the target area can be determined based on the coordinates of the projection point of the detection frame on the specified straight line and the coordinates of the calibration point.

[0128] For example, the location of a person in the target area may include, but is not limited to, the person's position relative to a preset calibration position, or the distance of the person from the front-end data acquisition device.

[0129] It can be seen that, in Figure 2 In the method flow shown, a calibration position for locating personnel is preset, as well as a straight line (i.e., a designated straight line) for assisting personnel location. The coordinates of the projection point (i.e., the calibration point) of the preset calibration position on the designated straight line are determined. For any personnel, the coordinates of the detection frame in the monitoring image of the current front-end acquisition device can be used to determine the coordinates of the projection point of the detection frame on the designated straight line. Then, the position of the personnel in the target area can be determined based on the coordinates of the projection point of the detection frame on the designated straight line and the coordinates of the calibration point. Personnel location is achieved by calibrating the position, which improves the intuitiveness of personnel location so that relevant personnel can more intuitively determine the position of the person to be located in the target area.

[0130] In some embodiments, the specified straight line is a straight line parallel to a specified coordinate axis, which is the x-axis or the y-axis;

[0131] The above-mentioned determination of the coordinates of the projection point of the detection box on a specified straight line based on the coordinate information of the detection box in the monitoring image of the current front-end acquisition device can include:

[0132] When the x-axis is specified as the coordinate axis, the x-coordinate of the left edge of the detection box is determined as the coordinate of the projection point of the detection box on the specified straight line;

[0133] or,

[0134] When the y-axis is specified as the coordinate axis, the y-coordinate of the lower edge of the detection box is determined as the coordinate of the projection point of the detection box on the specified straight line.

[0135] For example, let's take the y-axis as the specified coordinate axis. The implementation principle is similar when the x-axis is specified.

[0136] For any person, the coordinates of the projection point of the lower edge of the detection box on the specified straight line can be determined based on the detection box in the monitoring image of the current front-end acquisition device. That is, the coordinate y value of the lower edge of the detection box, and the coordinate y value of the lower edge of the detection box is determined as the coordinates of the projection point of the detection box on the specified straight line.

[0137] In one example, the aforementioned preset calibration location is the intersection of a preset positioning reference object and the road surface;

[0138] The determination of the person's position in the target area based on the coordinates of the projection point of the detection frame on the specified straight line and the coordinates of the calibration point may include:

[0139] If the coordinates of the projection point of the detection frame on the specified straight line are less than the coordinates of the calibration point corresponding to the first positioning reference, but greater than the coordinates of the calibration point corresponding to the second positioning reference, then the person is determined to be located between the first positioning reference and the second positioning reference; wherein, the first positioning reference and the second positioning reference are two preset positioning references adjacent to each other in the monitoring image of the current front-end acquisition device.

[0140] For example, in order to improve the intuitiveness of personnel positioning in the target area, positioning reference objects, such as street light poles or mounting brackets for video acquisition equipment, can be set in advance in the target area, and the intersection of the positioning reference object and the road surface can be determined as the calibration position. The coordinates of the projection points of each calibration position on the specified straight line can be determined in advance.

[0141] For example, when the specified coordinate axis is the y-axis, the coordinates of the projection points of each calibration position onto the specified straight line can be the y-coordinates of each calibration position.

[0142] Having determined the coordinates of the projection point of the detection box on the specified straight line in the manner described above, we can compare the coordinates of the projection point of the detection box on the specified straight line with the coordinates of the projection point of the calibration position on the specified straight line.

[0143] If the coordinates of the projection point of the detection frame on the specified straight line are less than the coordinates of the calibration point corresponding to the first positioning reference, but greater than the coordinates of the calibration point corresponding to the second positioning reference, then the person is determined to be located between the first and second positioning references.

[0144] For example, taking a pedestrian walkway on one side of a bridge as the target area, multiple streetlights are installed along the direction of traffic. These streetlights can be used as positioning references. For any front-end data acquisition device, the y-coordinate of the streetlight appearing in the monitoring screen of that device can be marked (i.e., using the aforementioned coordinate axis as the y-axis). For example, a schematic diagram could be as follows: Figure 3A As shown, the y-coordinates of each street light pole in the image can be marked. Therefore, based on the latest detection data of the object to be monitored and the y-coordinates of each street light pole in the image, the position of the object in the target area can be determined. For example, the detection data may include the detection frame position; the y-coordinates of the lower edge of the latest detection frame of the object to be monitored, and the y-coordinates of each street light pole in the image, can determine which two street light poles the object to be monitored is located between.

[0145] For example, once it is determined that the first person is located between those two lampposts, the relevant location can be notified to the relevant personnel based on the lamppost numbers and their actual positions, so that the relevant personnel can accurately locate the first person.

[0146] In some embodiments, the specified straight line is parallel to the optical axis of the current front-end acquisition device and passes through the projection point of the current front-end acquisition device on the road surface;

[0147] The above-mentioned determination of the coordinates of the projection point of the detection box on a specified straight line based on the coordinate information of the detection box in the monitoring image of the current front-end acquisition device can include:

[0148] Based on the coordinates of the center point of the detection frame and the coordinates of the calibration point, determine the coordinates of the projection point of the center point of the detection frame on the specified straight line;

[0149] The coordinates of the projection point of the center point of the detection frame onto the specified straight line are determined as the coordinates of the projection point of the detection frame onto the specified straight line.

[0150] For example, in order to locate a person in a target area, for any front-end acquisition device, on the one hand, a virtual line (i.e. the straight line specified above) can be set in the acquisition area of ​​the front-end acquisition device. The virtual line is parallel to the optical axis of the front-end acquisition device and passes through the projection point of the front-end acquisition device on the road surface.

[0151] For example, the projection point of the front-end acquisition device on the ground can be the intersection of the pole on which the front-end acquisition device is installed and the road surface.

[0152] It should be noted that the parallelism between the virtual line and the optical axis of the front-end acquisition device allows for tolerable errors.

[0153] On the other hand, multiple calibration positions can be set within the acquisition area of ​​the front-end acquisition device, and the coordinates of the projection points (i.e., calibration points) of the calibration positions on the virtual line can be calibrated (the coordinates of the calibration points in the monitoring screen of the front-end acquisition device).

[0154] Accordingly, for any person to be located, such as the first person determined to meet the conditions for staying in the monitoring area, the pixel coordinates of the projection point of the center point on the virtual line can be determined based on the pixel coordinates of the center point of the detection frame in the monitoring image of the current front-end acquisition device and the pixel coordinates of the calibration point.

[0155] like Figure 3B As shown, assuming the calibration points include points A, B, and C, their pixel coordinates are (x, y ... a y a ), (x b y b ), (x c y c The distances to the current data acquisition device are L, respectively. A LB L C The center pixel coordinates of the latest detection bounding box in the monitoring image of the current front-end acquisition device are (x'). d y' d Then, the pixel coordinates (x, y) of the projection point (which can be denoted as point D) of the center point on the virtual line can be determined by the following formula. d y d ):

[0156] k=(y a -y b ) / (x a -x b )

[0157] x d =(y' d -y b ) / k+x b

[0158] y d =y' d

[0159] In one example, determining the person's position in the target area based on the coordinates of the projection point of the detection box onto a specified straight line and the coordinates of the calibration point can include:

[0160] Based on the coordinates of the projection point of the center point of the detection frame on the specified straight line, the coordinates of each calibration point, and the physical distance from each calibration point to the current front-end acquisition device, determine the physical distance from the projection point of the center point on the specified straight line to the current front-end acquisition device.

[0161] The location of the person in the target area is determined based on the physical distance from the projection point of the center point on the specified straight line to the current front-end acquisition device.

[0162] For example, in order to locate personnel in a target area, the physical distance from each calibration point to the front-end data acquisition device can be preset.

[0163] The coordinates (x, y) of the projection point of the center point of the detection frame of the person to be located on the virtual line were determined in the above manner. d y d In the case of a center point projected onto a virtual line, the coordinates of each calibration point, and the physical distance from each calibration point to the current front-end acquisition device can be used to determine the physical distance (denoted as L) from the center point projected onto the virtual line to the current front-end acquisition device. D ).

[0164] For example, based on the principle that the cross ratio remains unchanged in the projection transformation of a straight line, we know that:

[0165]

[0166] Where AC is the physical distance between calibration point A and calibration point C, and the same applies to BC, BD, and AD. ac is the pixel distance between calibration point A and calibration point C in the current front-end acquisition device's image, and the same applies to bc, bd, and ad.

[0167] Assume the physical distance from point D to the current front-end data acquisition device is L. D Then AC = |L C -L A |;BC=|L C -L B |;BD=|L D -L B |;AD=|L D -L A |

[0168] Based on this, we can conclude that:

[0169]

[0170] Based on this, we can conclude that:

[0171]

[0172] Among them, ac, bc, bd, ad, L A L B L C All are known.

[0173] We can set:

[0174]

[0175] Therefore, we can obtain:

[0176]

[0177] or,

[0178]

[0179] L is determined based on the positional order of points A, B, C, and D. D The final value of .

[0180] Among them, sorted by y-coordinate from largest to smallest, if the y-coordinate of point D lies between the y-coordinates of two points A, B, and C, then L D It is also between the physical distances corresponding to these two points.

[0181] For example, if the y-coordinate of point D lies between the y-coordinates of points A and B, then LD The value of L is in A and L B .

[0182] If the y-value of point D is the minimum or maximum y-value among the y-values ​​of A, B, C, and D, then L D With L A L B L C In comparison, it is either the minimum or the maximum value.

[0183] For example, after determining the physical distance L between point D and the current front-end acquisition device... D In this case, it can be based on the physical distance L D Determine the location of the person to be located within the target area.

[0184] For example, this physical distance L can be... D The physical distance from the person to be located to the current front-end data acquisition device is determined (e.g., the physical distance from the person to the pole where the current data acquisition device is installed), and this is notified to the relevant personnel so that they can locate the first person.

[0185] To enable those skilled in the art to better understand the technical solutions provided in the embodiments of this application, the technical solutions provided in the embodiments of this application are described below with reference to specific examples.

[0186] In this embodiment, the target area is the pedestrian walkway on one side of the bridge.

[0187] For pedestrian walkways on either side of a bridge, multiple cameras can be installed at certain intervals for structured video analysis.

[0188] For example, cameras can be installed in pairs facing away from each other at a certain distance.

[0189] Among these multiple cameras, at least two adjacent cameras have a blind spot, meaning that the multiple cameras cannot cover the entire area of ​​the pedestrian walkway on that side.

[0190] The backend server can acquire videos captured by each camera. For any video stream from any camera, it can detect the optimal face and body, person ID, and face and body association ID every few seconds, and use certain image analysis algorithms to analyze the structured features of the face and the structured features of the body respectively.

[0191] For example, the optimal face can be the face with the highest score determined by a preset quality scoring algorithm; the optimal human body can be the human body with the highest score determined by a preset quality scoring algorithm.

[0192] It should be noted that, considering that there may be significant differences in the structural features of the same human body under different orientations, the optimal human body can include the optimal human body in different orientations, such as facing the camera, facing away from the camera, or facing the side of the camera.

[0193] For example, the Person ID is used to track the same person under the same camera. The distribution of capture points of the same person under the same camera can be determined based on the tracking algorithm, and a unique Person ID under that camera can be assigned to that person.

[0194] The Face-Body Association ID is used to associate the face and body of the same person. When the same person under the same camera is identified by the tracking algorithm, the same Face-Body Association ID can be configured for the structured facial features and structured body features of the same person.

[0195] For different cameras, structured feature clustering can be used to perform real-time face clustering on face images based on their structured facial features, assigning a unique ID to each person and completing their identity information. Similarly, real-time body clustering can be performed on body images based on their structured body features and person IDs. Furthermore, based on the associated IDs of faces and bodies, face and body clusters can be merged, thereby identifying the same person captured by different cameras and determining the distribution of capture points for the same person across different cameras.

[0196] For example, a unique person ID is used to uniquely identify a person (identifying the same person under different cameras).

[0197] For example, given the distribution of capture points for the same person under different cameras, the warning result can be determined based on this distribution and a preset warning algorithm strategy. The implementation process can be as follows: Figure 4 As shown.

[0198] For example, the preset early warning algorithm strategy may include, but is not limited to, a monitoring area stay early warning strategy and a blind area stay early warning strategy.

[0199] For example, if a person meets the conditions for a monitoring area stay warning based on the monitoring area stay warning strategy, such as when the number of times a person is captured by the same front-end acquisition device exceeds a first preset number, and the time when the first person is continuously captured by the front-end acquisition device reaches a first preset time, a suspected monitoring area stay warning can be issued.

[0200] Based on the blind spot loitering warning strategy, if a person meets the blind spot loitering warning conditions, for example, if no data is captured on the person by the next front-end acquisition device for a continuous second time period, a suspected blind spot loitering warning can be issued.

[0201] like Figure 5 As shown, for any person, if the distribution of capture points for that person in the current monitoring area (i.e., the pedestrian passage on one side of the bridge) is determined in accordance with the method described in the above embodiments, the number of times that person is captured by the current camera and the duration of continuous capture by the current camera can be counted in any capture of that person.

[0202] If a person is captured by the current camera more than a first preset number of times, and the duration of continuous capture by the current camera on that person reaches a first preset threshold, the system can use the person detection frame and the coordinates marked on the streetlights deployed on that side of the pedestrian walkway (see [reference]). Figure 3A It outputs warnings of suspected lingering in the monitoring area and the target location of personnel.

[0203] If it is determined that the person does not meet the conditions for staying in the monitored area, the capture event can be recorded (the recorded information may include, but is not limited to, the capture time, the position of the person detection frame, etc.), and we can wait for subsequent capture events of the person.

[0204] If the location of the person's camera is determined based on the distribution of the camera's capture points, the person's record can be deleted.

[0205] If the distribution of the person's capture points determines that the person has entered a monitoring blind zone, it can be determined whether the person meets the conditions for staying in the blind zone. If the conditions are met, the corresponding monitoring blind zone can be obtained based on the last capture data of the person in the monitoring area, and a suspected blind zone stay warning and the corresponding monitoring blind zone location can be output.

[0206] For example, for any individual, once the blind spot of that individual is identified, preset positions for each blind spot can be set on a high-altitude PTZ camera, and the corresponding preset position on the PTZ camera can be linked to verify whether the person is in the blind spot. If a warning of a suspected loitering individual is issued, measures such as calling to persuade them to leave, notifying nearby personnel to check and confirm, and contacting relatives to confirm their departure can be taken through checking the file of the suspected loitering individual.

[0207] The method provided in this application has been described above. The apparatus provided in this application is described below:

[0208] Please see Figure 6 This is a schematic diagram of the structure of a monitoring device provided in an embodiment of this application, as shown below. Figure 6 As shown, the monitoring device may include:

[0209] The acquisition unit 610 is configured to acquire image data collected by front-end acquisition devices deployed in a target area; wherein, multiple front-end acquisition devices are deployed in the target area; and the acquisition area of ​​each front-end acquisition device is a monitoring area.

[0210] The query unit 620 is configured to, for any first person to be monitored, query the image data based on the captured data of the first person in the current monitoring area; the current monitoring area is the acquisition area of ​​the current front-end acquisition device that captured the first person; the first person is any person identified after performing personnel clustering on the image data;

[0211] Analysis unit 630 is configured to determine, based on the captured data, whether the first person meets the conditions for staying in the monitored area;

[0212] The processing unit 640 is configured to issue a first warning if the first person meets the conditions for staying in the monitoring area; and to determine whether to issue a warning based on the subsequent capture data of the first person if the first person does not meet the conditions for staying in the monitoring area.

[0213] In some embodiments, the processing unit 640 is further configured to, if the first person meets the conditions for staying in the monitoring area, determine the coordinates of the projection point of the detection frame on a specified straight line based on the coordinate information of the detection frame in the monitoring image of the current front-end acquisition device; and determine the position of the first person in the target area based on the coordinates of the projection point of the detection frame on the specified straight line and the coordinates of the calibration point; wherein the calibration point is the projection point of the preset calibration position on the specified straight line.

[0214] In some embodiments, the specified line is a line parallel to a specified coordinate axis, which is the x-axis or the y-axis;

[0215] The processing unit 640 is specifically configured to, when the specified coordinate axis is the x-axis, determine the x-coordinate of the left edge of the detection box as the coordinate of the projection point of the detection box on the specified straight line; or, when the specified coordinate axis is the y-axis, determine the y-coordinate of the lower edge of the detection box as the coordinate of the projection point of the detection box on the specified straight line.

[0216] In some embodiments, the preset calibration position is the intersection of a preset positioning reference and the road surface.

[0217] The processing unit 640 is specifically configured to determine that the first person is located between the first positioning reference and the second positioning reference when the coordinates of the projection point of the detection frame on the specified straight line are less than the coordinates of the calibration point corresponding to the first positioning reference and greater than the coordinates of the calibration point corresponding to the second positioning reference; wherein the first positioning reference and the second positioning reference are two preset positioning references adjacent to the monitoring image of the current front-end acquisition device.

[0218] In some embodiments, the specified straight line is parallel to the optical axis of the current front-end acquisition device and passes through the projection point of the current front-end acquisition device on the road surface;

[0219] The processing unit 640 is specifically configured to determine the coordinates of the projection point of the center point of the detection frame on a specified straight line based on the coordinates of the center point of the detection frame and the coordinates of the calibration point; and to determine the coordinates of the projection point of the center point of the detection frame on the specified straight line as the coordinates of the projection point of the detection frame on the specified straight line.

[0220] In some embodiments, the processing unit 640 is specifically configured to determine the physical distance from the projection point of the center point on the specified straight line to the current front-end acquisition device based on the coordinates of the projection point of the center point of the detection frame on the specified straight line, the coordinates of each calibration point, and the physical distance from each calibration point to the current front-end acquisition device; and to determine the position of the first person in the target area based on the physical distance from the projection point of the center point on the specified straight line to the current front-end acquisition device.

[0221] In some embodiments, the processing unit 640 is specifically configured to issue a suspected blind spot lingering warning for the first person if no capture data of the first person is received from the next front-end acquisition device for a continuous second time period; the suspected blind spot lingering warning is used to indicate that the first person has entered the monitoring blind spot between the current front-end acquisition device and the next front-end acquisition device.

[0222] In some embodiments, the processing unit 640 is further configured to link a PTZ camera associated with the blind spot where the first person is located to acquire an image of the blind spot, and determine whether the first person is in the blind spot based on the image.

[0223] Please see Figure 7 This is a schematic diagram of the structure of a personnel positioning device provided in an embodiment of this application, as shown below. Figure 7 As shown, the personnel positioning device may include

[0224] The acquisition unit 710 is configured to acquire image data collected by a front-end acquisition device deployed in the target area;

[0225] The determining unit 720 is configured to, for any person, determine the coordinates of the projection point of the detection box on a specified straight line based on the coordinate information of the detection box in the monitoring image of the current front-end acquisition device;

[0226] The positioning unit 730 is configured to determine the position of the person in the target area based on the coordinates of the projection point of the detection frame on a specified straight line and the coordinates of the calibration point; wherein the calibration point is the projection point of a preset calibration position on the specified straight line.

[0227] In some embodiments, the specified straight line is a straight line parallel to a specified coordinate axis, which is the x-axis or the y-axis;

[0228] The determining unit 720 is specifically configured to, when the specified coordinate axis is the x-axis, determine the x-coordinate of the left edge of the detection box as the coordinate of the projection point of the detection box on the specified straight line; or, when the specified coordinate axis is the y-axis, determine the y-coordinate of the lower edge of the detection box as the coordinate of the projection point of the detection box on the specified straight line.

[0229] In some embodiments, the preset calibration position is the intersection of a preset positioning reference object and the road surface;

[0230] The positioning unit 730 is specifically configured to determine that the person is located between the first positioning reference and the second positioning reference when the coordinates of the projection point of the detection frame on the specified straight line are less than the coordinates of the calibration point corresponding to the first positioning reference and greater than the coordinates of the calibration point corresponding to the second positioning reference; wherein the first positioning reference and the second positioning reference are two preset positioning references adjacent to the monitoring image of the current front-end acquisition device.

[0231] In some embodiments, the specified straight line is parallel to the optical axis of the current front-end acquisition device and passes through the projection point of the current front-end acquisition device on the road surface;

[0232] The determining unit 720 is specifically configured to determine the coordinates of the projection point of the center point of the detection frame on a specified straight line based on the coordinates of the center point of the detection frame and the coordinates of the calibration point; and to determine the coordinates of the projection point of the center point of the detection frame on the specified straight line as the coordinates of the projection point of the detection frame on the specified straight line.

[0233] In some embodiments, the positioning unit 730 is specifically configured to determine the physical distance from the projection point of the center point on the specified straight line to the current front-end acquisition device based on the coordinates of the projection point of the center point of the detection frame on the specified straight line, the coordinates of each calibration point, and the physical distance from each calibration point to the current front-end acquisition device; and to determine the position of the person in the target area based on the physical distance from the projection point of the center point on the specified straight line to the current front-end acquisition device.

[0234] Please see Figure 8This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. The electronic device may include a processor 801, a communication interface 802, a memory 803, and a communication bus 804. The processor 801, communication interface 802, and memory 803 communicate with each other via the communication bus 804. The memory 803 stores a computer program; the processor 801 can execute the monitoring or personnel positioning methods described above by executing the program stored in the memory 803.

[0235] The memory 803 mentioned in this document can be any electronic, magnetic, optical, or other physical storage device that can contain or store information, such as executable instructions, data, etc. For example, the memory 803 can be: RAM (Random Access Memory), volatile memory, non-volatile memory, flash memory, storage drive (such as hard disk drive), solid-state drive, any type of storage disk (such as optical disc, DVD, etc.), or similar storage media, or combinations thereof.

[0236] This application also provides a non-transitory machine-readable storage medium storing a computer program, such as... Figure 8 The memory 803 in the computer program can be used by Figure 8 The processor 801 in the electronic device shown executes to implement the monitoring or personnel location method described above.

[0237] This application also provides a computer program stored on a non-transitory machine-readable storage medium, such as... Figure 8 The memory 803 in the processor, and when the processor executes the computer program, it causes the processor 801 to perform the monitoring or personnel location method described above.

[0238] This application also provides a monitoring system, including: Figure 8 The electronic device shown may further include a front-end acquisition device and a personnel clustering analysis device. The front-end acquisition device is used to acquire image data of personnel and send it to the personnel clustering analysis device. The personnel clustering analysis device is used to perform personnel clustering on the image data to identify image data of the same person and send the identification results to the electronic device. The electronic device can implement monitoring in the manner described in the above embodiments.

[0239] In some embodiments, the monitoring system provided in this application may further include a PTZ camera; the PTZ camera is provided with at least one preset position, each preset position is associated with at least one monitoring blind zone, each monitoring blind zone is a monitoring blind zone between two adjacent front-end acquisition devices, and the PTZ camera is used to adjust to the position corresponding to the preset position indicated by the instruction according to the instruction, and to acquire image data of the monitoring blind zone associated with the preset position.

[0240] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0241] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for locating personnel, characterized in that, include: Acquire image data collected by front-end acquisition devices deployed in the target area; For any person, based on the pixel coordinates of the detection box in the monitoring image of the current front-end acquisition device, determine the pixel coordinates of the projection point of the detection box on a specified straight line; the specified straight line is a preset straight line used to assist in the positioning of the person; the specified straight line is located in the monitoring image; The position of the person in the target area is determined based on the pixel coordinates of the projection point of the detection frame on the specified straight line and the pixel coordinates of the calibration point; wherein, the calibration point is the projection point of the preset calibration position on the specified straight line; the position of the person in the target area includes the position of the person relative to the preset calibration position, or the distance of the person from the front-end acquisition device; The distance between the personnel and the front-end data collection device is determined in the following way: Based on the pixel coordinates of the projection point of the detection frame on the specified straight line, the pixel coordinates of each calibration point, and the physical distance from each calibration point to the current front-end acquisition device, determine the physical distance from the projection point of the detection frame on the specified straight line to the current front-end acquisition device. The location of the person in the target area is determined based on the physical distance from the projection point of the detection frame on the specified straight line to the current front-end acquisition device.

2. The method according to claim 1, characterized in that, The specified straight line is a straight line parallel to a specified coordinate axis, which is either the x-axis or the y-axis. The step of determining the pixel coordinates of the projection point of the detection box on a specified straight line based on the pixel coordinate information of the detection box in the monitoring image of the current front-end acquisition device includes: When the specified coordinate axis is the x-axis, the pixel coordinate x value of the left edge of the detection box is determined as the pixel coordinate of the projection point of the detection box on the specified straight line; or, When the specified coordinate axis is the y-axis, the pixel coordinate y-value of the lower edge of the detection box is determined as the pixel coordinate of the projection point of the detection box on the specified straight line.

3. The method according to claim 2, characterized in that, The preset calibration position is the intersection of a preset positioning reference object and the road surface; Determining the position of the person in the target area based on the pixel coordinates of the projection point of the detection frame on the specified straight line and the pixel coordinates of the calibration point includes: If the pixel coordinates of the projection point of the detection frame on the specified straight line are less than the pixel coordinates of the calibration point corresponding to the first positioning reference, but greater than the pixel coordinates of the calibration point corresponding to the second positioning reference, then the person is determined to be located between the first positioning reference and the second positioning reference; wherein, the first positioning reference and the second positioning reference are two preset positioning references adjacent to each other in the monitoring image of the current front-end acquisition device.

4. The method according to claim 1, characterized in that, The specified straight line is parallel to the optical axis of the current front-end acquisition device and passes through the projection point of the current front-end acquisition device on the road surface; The step of determining the pixel coordinates of the projection point of the detection box on a specified straight line based on the pixel coordinate information of the detection box in the monitoring image of the current front-end acquisition device includes: Based on the pixel coordinates of the center point of the detection frame and the pixel coordinates of the calibration point, determine the pixel coordinates of the projection point of the center point of the detection frame on the specified straight line; The pixel coordinates of the projection point of the center point of the detection box onto the specified straight line are determined as the pixel coordinates of the projection point of the detection box onto the specified straight line.

5. A personnel positioning device, characterized in that, include: The acquisition unit is configured to acquire image data collected by front-end acquisition devices deployed in the target area; The determining unit is configured to, for any person, determine the pixel coordinates of the projection point of the detection box on a specified straight line based on the pixel coordinate information of the detection box in the monitoring image of the current front-end acquisition device; the specified straight line is a preset straight line used to assist in the positioning of the person; the specified straight line is located in the monitoring image; The positioning unit is configured to determine the position of the person in the target area based on the pixel coordinates of the projection point of the detection frame on a specified straight line and the pixel coordinates of the calibration point; wherein the calibration point is the projection point of a preset calibration position on the specified straight line; the position of the person in the target area includes the position of the person relative to the preset calibration position, or the distance of the person from the front-end acquisition device; Specifically, the positioning unit is configured to determine the physical distance from the projection point of the detection frame on the specified straight line to the current front-end acquisition device based on the pixel coordinates of the projection point of the detection frame on the specified straight line, the pixel coordinates of each calibration point, and the physical distance from each calibration point to the current front-end acquisition device; and to determine the position of the person in the target area based on the physical distance from the projection point of the detection frame on the specified straight line to the current front-end acquisition device.

6. The apparatus according to claim 5, characterized in that, The specified straight line is a straight line parallel to a specified coordinate axis, which is either the x-axis or the y-axis. The determining unit is specifically configured to, when the specified coordinate axis is the x-axis, determine the pixel coordinate x value of the left edge of the detection box as the pixel coordinate of the projection point of the detection box on the specified straight line; or, when the specified coordinate axis is the y-axis, determine the pixel coordinate y value of the lower edge of the detection box as the pixel coordinate of the projection point of the detection box on the specified straight line. The preset calibration position is the intersection of a preset positioning reference object and the road surface; The positioning unit is specifically configured to determine that the person is located between the first positioning reference and the second positioning reference when the pixel coordinates of the projection point of the detection frame on the specified straight line are less than the pixel coordinates of the calibration point corresponding to the first positioning reference and greater than the pixel coordinates of the calibration point corresponding to the second positioning reference; wherein the first positioning reference and the second positioning reference are two preset positioning references adjacent to each other in the monitoring image of the current front-end acquisition device.

7. The apparatus according to claim 5, characterized in that, The specified straight line is parallel to the optical axis of the current front-end acquisition device and passes through the projection point of the current front-end acquisition device on the road surface; The determining unit is specifically configured to determine the pixel coordinates of the projection point of the center point of the detection frame on the specified straight line based on the pixel coordinates of the center point of the detection frame and the pixel coordinates of the calibration point; and to determine the pixel coordinates of the projection point of the center point of the detection frame on the specified straight line as the pixel coordinates of the projection point of the detection frame on the specified straight line.

8. An electronic device, characterized in that, include: The processor, communication interface, memory, and communication bus are connected, with the processor, communication interface, and memory communicating with each other via the communication bus. Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the method as described in any one of claims 1 to 4.

9. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 4.

Images