WO2022135242A1

WO2022135242A1 - Virtual positioning method and apparatus, and virtual positioning system

Info

Publication number: WO2022135242A1
Application number: PCT/CN2021/138525
Authority: WO
Inventors: 聂兰龙
Original assignee: 青岛千眼飞凤信息技术有限公司
Priority date: 2020-12-21
Filing date: 2021-12-15
Publication date: 2022-06-30
Also published as: CN114648572A

Abstract

Disclosed are a virtual positioning method and apparatus, and a virtual positioning system. The method comprises: acquiring target positioning data of a target in a first camera cluster, wherein the first camera cluster is a camera which is used in a positioning system, and the positioning system is used for executing a target tracking task in order to generate positioning data; sending the target positioning data to the positioning system, wherein the positioning system generates virtual positioning data of the target in a second camera cluster by using a target mapping model and the target positioning data, and the target mapping model is used for describing a mapping relationship between the first camera cluster and the second camera cluster in a space; and performing virtual positioning on the target according to the virtual positioning data. By means of the present application, the technical problems in the relevant art of the application of an image sensor being relatively simple, and it not being possible to perform large-scale collaborative tracking by using the image sensor are solved.

Description

Virtual positioning method and device, virtual positioning system

This application claims the priority of the Chinese patent application with the application number 202011522213.2 and the application title "Virtual Positioning Method and Device, Virtual Positioning System" filed with the China Patent Office on December 21, 2020, the entire contents of which are incorporated into this application by reference middle.

technical field

The present application relates to the field of computer technology, and in particular, to a virtual positioning method and device, and a virtual positioning system.

Background technique

With the exponential development of technology, the production cost of image sensors will decrease year by year, and the production scale will also increase exponentially. At the same time, various new applications will be born, such as chain image sensor technology. As image sensors are widely used in social scenes, a technology for coordinated tracking of multiple image sensors is also required to meet the user's target tracking needs.

At present, the use of image sensors for positioning and tracking technology has been widely used, such as some excellent tracking systems PTAM (Parallel Tracking and Mapping), ACTS (Automatic Camera Tracking System) and so on. Different positioning and tracking technologies have different advantages and maturity, but these technologies are all applied to identification, positioning and tracking under a single surveillance camera.

Currently, there is a lack of a large-scale coordinated tracking technology for image sensors.

Aiming at the problem that the application of the image sensor in the above-mentioned related art is relatively simple, and the image sensor cannot be used for large-scale collaborative tracking, no effective solution has been proposed yet.

Application content

Embodiments of the present application provide a virtual positioning method and device, and a virtual positioning system, so as to at least solve the technical problem in the related art that the application of image sensors is relatively simple and large-scale collaborative tracking cannot be performed by using image sensors.

According to an aspect of the embodiments of the present application, a virtual positioning method is provided, including: acquiring target positioning data of a target in a first camera cluster, where the first camera cluster is a camera used in a positioning system, The positioning system is used to perform a target tracking task to generate positioning data; the target positioning data is sent to the positioning system, wherein the positioning system utilizes the target mapping model and the target positioning data to generate the target at Virtual positioning data in the second camera cluster, the target mapping model is used to describe the spatial mapping relationship between the first camera cluster and the second camera cluster; virtual positioning of the target is performed according to the virtual positioning data position.

Optionally, the target mapping model is a mapping model obtained by the positioning system from a mapping model system, wherein the mapping model system generates a relationship between the positions of the cameras and the viewing angle in the predetermined application scene by initializing the modeling. the mapping model.

Optionally, before performing virtual positioning on the target according to the virtual positioning data, the virtual positioning method further includes: verifying the virtual positioning data; wherein, verifying the virtual positioning data includes: : obtain the actual positioning data of the target in the second camera cluster; determine the similarity between the actual positioning data and the virtual positioning data by using a predetermined verification rule, so as to compare the virtual positioning data with the actual positioning data. Check the consistency before positioning the data.

Optionally, the virtual positioning method further includes: in response to a relay tracking request signal, and determining the best sampling camera in the second camera cluster according to the virtual positioning data; setting the attribute of the best sampling camera as obtaining the updated first camera cluster from the first camera cluster; acquiring the image information stream collected by the updated first camera cluster; sending the image information stream to the positioning system, wherein the positioning system is based on The image information stream generates target positioning data for the target in the updated first camera cluster.

Optionally, in response to the relay tracking request signal, the method includes: determining the distribution density of cameras in a predetermined application scenario where the first camera cluster and the second camera cluster are located; when it is determined that the distribution density is less than a predetermined value, in the When the target is located at a predetermined position on the edge of the viewing area of the first camera cluster, in response to the relay tracking request signal; or, when it is determined that the distribution density is not less than a predetermined value, when the target leaves the first camera cluster for viewing in the middle of the zone, in response to the relay tracking request signal.

Optionally, responding to the relay tracking request signal includes: determining that the positioning accuracy of the target is lower than a predetermined threshold; and responding to the relay tracking request signal.

Optionally, after performing virtual positioning on the target according to the virtual positioning data, the virtual positioning method further includes: in response to an identification matching task, determining that the second camera cluster includes at least one camera of the target. obtaining frame image information of the at least one camera; identifying a feature identifier in the frame image information, and generating target identifier information of the target based on the feature identifier.

Optionally, performing virtual positioning on the target according to the virtual positioning data includes: responding to a target query instruction of a terminal device, and acquiring target identification information of the target query instruction; combining the target identification information based on the target identification information The target positioning data and/or the virtual positioning data are used to determine a viewfinder camera; and the video stream of the target collected by the viewfinder camera is fed back to the terminal device.

Optionally, before performing virtual positioning on the target according to the virtual positioning data, the virtual positioning method further includes: in response to a tracking task request, determining a positioning calibration in the second camera cluster according to the virtual positioning data camera; acquiring frame image buffer information of the positioning calibration camera; generating calibration positioning data of the target in the positioning calibration camera based on the frame image buffer information; correcting the target positioning data according to the calibration positioning data.

Optionally, the start condition of the tracking task request includes one of the following: the target is lost, the determination accuracy of the target is lower than a predetermined threshold, and the predetermined plan of the target tracking task.

Optionally, the virtual positioning method further includes: determining a viewfinder camera according to a predetermined rule according to the target positioning data and/or the virtual positioning data; and sending frame image information collected by the viewfinder camera to a predetermined storage medium.

According to another aspect of the embodiments of the present application, a virtual positioning device is also provided, including: an obtaining unit configured to obtain target positioning data of a target in a first camera cluster, wherein the first camera cluster is a positioning system The camera used in the positioning system is used to perform a target tracking task to generate positioning data; a sending unit is configured to send the target positioning data to the positioning system, wherein the positioning system utilizes target mapping The model and the target positioning data generate virtual positioning data of the target in the second camera cluster, and the target mapping model is used to describe the spatial mapping relationship between the first camera cluster and the second camera cluster; A virtual positioning unit, configured to perform virtual positioning on the target according to the virtual positioning data.

Optionally, the virtual positioning device further includes: a verification unit, configured to verify the virtual positioning data before performing virtual positioning on the target according to the virtual positioning data; wherein, the verification unit , comprising: a first acquisition module, configured to acquire the actual positioning data of the target in the second camera cluster; a verification module, configured to use a predetermined verification rule to determine the actual positioning data and the virtual positioning data to verify the consistency between the virtual positioning data and the actual positioning data.

Optionally, the virtual positioning device further includes: a first determining unit configured to respond to a relay tracking request signal and determine the best sampling camera in the second camera cluster according to the virtual positioning data; a setting unit configured to set In order to set the attribute of the best sampling camera as the first camera cluster, the updated first camera cluster is obtained; the acquiring unit is configured to acquire the image information stream collected by the updated first camera cluster; The sending unit is configured to send the image information stream to the positioning system, wherein the positioning system generates target positioning data of the target in the updated first camera cluster based on the image information stream .

Optionally, the first determination unit includes: a first determination module, configured to determine the distribution density of cameras in the predetermined application scenarios where the first camera cluster and the second camera cluster are located; a second determination module, configured to determine In order to determine that the distribution density is less than a predetermined value, when the target is located at a predetermined position on the edge of the viewing area of the first camera cluster, in response to the relay tracking request signal; or, the third determining module is configured to determine the When the distribution density is not less than a predetermined value, when the target leaves the middle position of the viewing area of the first camera cluster, it responds to the relay tracking request signal.

Optionally, the first determination unit includes: a fourth determination module configured to determine that the positioning accuracy of the target is lower than a predetermined threshold; and a response module configured to respond to the relay tracking request signal.

Optionally, the virtual positioning device further includes: a second determining unit, configured to, after virtual positioning of the target according to the virtual positioning data, in response to an identification matching task, determine that the second camera cluster contains at least one camera of the target; the acquiring unit is configured to acquire the frame image information of the at least one camera; the first generating unit is configured to recognize the feature identifier in the frame image information, and based on the feature identifier Target identification information for the target is generated.

Optionally, the virtual positioning unit includes: a second acquisition module, configured to respond to a target query instruction of a terminal device, and acquire target identification information of the target query instruction; a fifth determination module, set to be based on the The target identification information is combined with the target positioning data and/or the virtual positioning data to determine a viewfinder camera; the feedback module is configured to feed back the video stream of the target collected by the viewfinder camera to the terminal device.

Optionally, the virtual positioning device further includes: a third determining unit, configured to, in response to a tracking task request, determine the third determining unit according to the virtual positioning data before performing virtual positioning on the target according to the virtual positioning data. A positioning calibration camera in a two-camera cluster; the obtaining unit is configured to obtain the frame image cache information of the positioning calibration camera; the second generating unit is configured to generate the target at the positioning based on the frame image cache information The calibration positioning data of the camera is calibrated; the correction unit is configured to correct the target positioning data according to the calibration positioning data.

Optionally, the virtual positioning device further includes: a fourth determining unit, configured to determine a viewfinder camera according to a predetermined rule according to the target positioning data and/or the virtual positioning data; the sending unit, configured to The frame image information collected by the viewfinder camera is sent to a predetermined storage medium.

According to another aspect of the embodiments of the present application, a virtual positioning system is also provided, using any one of the virtual positioning methods described above.

According to another aspect of the embodiments of the present application, a computer-readable storage medium is also provided, the computer-readable storage medium includes a stored computer program, wherein when the computer program is run by a processor, the computer is controlled The device where the storage medium is located executes the virtual positioning method described in any one of the above.

According to another aspect of the embodiments of the present application, a processor is also provided, where the processor is configured to run a computer program, wherein when the computer program runs, the virtual positioning method described in any one of the above is executed.

In the embodiment of the present application, the target positioning data of the target in the first camera cluster is obtained, wherein the first camera cluster is the camera used in the positioning system, and the positioning system is used to perform the target tracking task to generate positioning data. Send the target positioning data to the positioning system, wherein the positioning system uses the target mapping model and the target positioning data to generate virtual positioning data of the target in the second camera cluster, and the target mapping model is used to describe the first camera cluster and the second camera cluster The mapping relationship in space; the virtual positioning of the target is carried out according to the virtual positioning data, and the virtual positioning method provided by the embodiment of the present application realizes the purpose of positioning the target through the coordination of the camera cluster, and improves the accuracy of positioning the target. The technical effect of high degree is solved, and the technical problem that the application of the image sensor in the related art is relatively single, and the large-scale collaborative tracking cannot be carried out by using the image sensor.

Description of drawings

The drawings described herein are used to provide further understanding of the present application and constitute a part of the present application. The schematic embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation of the present application. In the attached image:

1 is a flowchart of a virtual positioning method according to an embodiment of the present application;

2 is a sequence diagram 1 of a virtual positioning method according to an embodiment of the present application;

3 is a second sequence diagram of a virtual positioning method according to an embodiment of the present application;

4 is a sequence diagram 3 of a virtual positioning method according to an embodiment of the present application;

5 is a sequence diagram 4 of a virtual positioning method according to an embodiment of the present application;

6 is a sequence diagram 5 of a virtual positioning method according to an embodiment of the present application;

7 is a sequence diagram 6 of a virtual positioning method according to an embodiment of the present application;

8 is a sequence diagram 7 of a virtual positioning method according to an embodiment of the present application;

9 is a sequence diagram 8 of a virtual positioning method according to an embodiment of the present application;

Figure 10 (a) is a schematic diagram 1 of a scenic spot according to an embodiment of the present application;

Figure 10 (b) is a schematic diagram 2 of a scenic spot according to an embodiment of the present application;

11 is a schematic diagram of a logistics transfer station according to an embodiment of the present application;

12 is a schematic diagram of a traffic monitoring area according to an embodiment of the present application;

13 is a schematic diagram of a kindergarten according to an embodiment of the present application;

14 is a schematic diagram of a subway operating company according to an embodiment of the present application;

FIG. 15 is a schematic diagram of a virtual positioning apparatus according to an embodiment of the present application.

Detailed ways

In order to make those skilled in the art better understand the solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only The embodiments are part of the present application, but not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the scope of protection of the present application.

It should be noted that the terms "first", "second", etc. in the description and claims of the present application and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific sequence or sequence. It is to be understood that data so used may be interchanged under appropriate circumstances so that the embodiments of the application described herein can be practiced in sequences other than those illustrated or described herein. Furthermore, the terms "comprising" and "having" and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those expressly listed Rather, those steps or units may include other steps or units not expressly listed or inherent to these processes, methods, products or devices.

According to an embodiment of the present application, a method embodiment of a virtual positioning method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings may be executed in a computer system such as a set of computer-executable instructions, and, Although a logical order is shown in the flowcharts, in some cases steps shown or described may be performed in an order different from that herein.

Fig. 1 is the flow chart of the virtual positioning method according to the embodiment of the present application, as shown in Fig. 1, this virtual positioning method comprises the following steps:

Step S102, acquiring target positioning data of the target in the first camera cluster, where the first camera cluster is a camera used in the positioning system, and the positioning system is used to perform a target tracking task to generate positioning data.

Optionally, the first camera cluster here is a sampling camera of a tracking and positioning system (ie, a positioning system), and the tracking and positioning system is configured to perform a target tracking task to generate positioning data.

Step S104, sending the target positioning data to the positioning system, wherein the positioning system uses the target mapping model and the target positioning data to generate virtual positioning data of the target in the second camera cluster, and the target mapping model is used to describe the first camera cluster and the second camera cluster. The mapping relationship of camera clusters in space.

Optionally, the cameras in the second camera cluster and the cameras in the first camera cluster are fixed cameras, the position and angle of the cameras are fixed, and frame image information with a still background can be obtained; The angle is fixed, and the zooming of the camera will not change the perspective relationship of image acquisition, but will only change the viewing area. Similarly, as long as the camera is preset with zoom parameters, the position coordinates of the target in various zoom states can be calculated.

In addition, the lens distortion problem occurs in the zooming process of some cameras, especially the wide-angle camera and fisheye camera have obvious distortion, and the effect of zooming on the target positioning can be eliminated through the distortion correction technology. Therefore, on the premise that the camera zoom parameters, lens distortion parameters and the modeling initialization focal length state are known, the target positioning data and the virtual positioning data are not affected by the camera zoom factor and the lens distortion factor.

It should be noted that, in the embodiment of the present application, the second camera cluster may also include a chain image acquisition device, and a plurality of cameras of the chain image acquisition device are distributed in a chain in the same data transmission bus.

In addition, in this embodiment of the present application, the tracking and positioning system may include one or more sets of target tracking algorithms, such as frame difference method, background compensation method, expectation maximization method, optical flow method, statistical model method, level set method and Parallel tracking and drawing PTAM (Parallel Tracking and Mapping, referred to as PTAM), automatic camera tracking system ACTS (Automatic Camera Tracking System, referred to as ACTS), etc., can also include deep neural networks, through machine learning and other means to achieve target detection Track and locate.

Step S106, perform virtual positioning on the target according to the virtual positioning data.

As can be seen from the above, in the embodiment of the present application, the target positioning data of the target in the first camera cluster can be obtained, wherein the first camera cluster is the camera used in the positioning system, and the positioning system is used to perform the target tracking task. to generate positioning data; send the target positioning data to the positioning system, wherein the positioning system uses the target mapping model and the target positioning data to generate virtual positioning data of the target in the second camera cluster, and the target mapping model is used to describe the first camera cluster and The mapping relationship of the second camera cluster in space; the virtual positioning of the target according to the virtual positioning data, the goal of coordinating the positioning of the target through the camera cluster is achieved, and the technical effect of improving the accuracy of the positioning of the target is achieved.

Therefore, the virtual positioning method provided by the embodiments of the present application solves the technical problem that the application of the image sensor in the related art is relatively simple, and the image sensor cannot be used for large-scale collaborative tracking.

In an optional embodiment, the target mapping model is a mapping model obtained by the positioning system from the mapping model system, wherein the mapping model system generates a mapping model including the relationship between the positions of each camera and the viewing angle in the predetermined application scene by means of initializing modeling the mapping model.

In this embodiment, the camera cluster mapping model is a space model generated by the mapping model system through initialization modeling and includes the relationship between the positions of each camera in the application scene and the viewing angle, and describes the first camera cluster and the second camera cluster in the space. mutual mapping relationship.

It should be noted that, in the embodiment of the present application, the mapping model may be one of a two-dimensional model, a three-dimensional model, and a dynamic model; wherein, the two-dimensional model is basically obtained by analyzing the image information of the initialization modeling frame collected by each camera. The overlapping area is generated, and the two-dimensional model has the problems of low positioning accuracy and easy to be affected by the three-dimensional environment in the actual application process. It is only suitable for application scenarios where the positioning accuracy is not high; among them, the three-dimensional model can be analyzed by The three or more common position points in the initialized modeling frame image information collected by the camera during the modeling initialization process are used to establish the three-dimensional function mapping relationship of each camera in the space model; the three-dimensional model can also be analyzed by analyzing a target in the modeling initialization process. Modeling is performed using the method of position coordinates at different times; since the camera clusters in the embodiments of the present application are all fixed positions and fixed viewing angles, the background of each camera viewing area is also the same at different times, and the same target moves to more than three The three-dimensional function mapping relationship of each camera in the space model can be constructed by analyzing the position coordinates of the target position in the framing area of each camera; among them, the dynamic model is the different space models corresponding to different conditions, and the position that can be standardized Repeated correspondence is a prerequisite for building a dynamic model. For example, subway trains will accurately stop at the same location when they arrive at the station. The spatial mapping relationship between the cameras inside and outside the doors of subway trains is a prerequisite for the train to accurately park on the platform. For another example, when an elevator stops on each floor, the location can be repeated and standardized. After a dynamic model is constructed, different 3D models can be triggered and determined according to different elevator parking conditions, and the 2D or 3D mapping relationship of different camera clusters can be obtained.

FIG. 2 is a sequence diagram 1 of a virtual positioning method according to an embodiment of the present application. As shown in FIG. 2 , the first camera cluster will send the collected image frame information stream to the service system, and the service system will send the frames of the first camera cluster to the service system. The image information stream is sent to the tracking and positioning system. The tracking and positioning system generates positioning data based on the frame image information flow, and returns the positioning data to the service system. The service system will send the positioning data of the target in the first camera cluster to the virtual positioning system. The positioning system will send a mapping model request to the mapping model system, and receive the target mapping model that the mapping model system requests feedback based on the mapping model. Virtual positioning data, and feedback the virtual positioning data to the service system.

That is, in this embodiment, the camera cluster mapping model can be extracted by acquiring the positioning data of the target in the first camera cluster; and virtual positioning data corresponding to the target in the second camera cluster can be generated based on the mapping model and the target positioning data.

In an optional embodiment, before performing virtual positioning on the target according to the virtual positioning data, the virtual positioning method may further include: verifying the virtual positioning data; wherein, verifying the virtual positioning data includes: Obtain the actual positioning data of the target in the second camera cluster; determine the similarity between the actual positioning data and the virtual positioning data by using a predetermined verification rule, so as to verify the consistency between the virtual positioning data and the actual positioning data.

In this embodiment, the positioning data in the second camera cluster is target positioning data generated by a second tracking system, and the second target positioning system is configured to sample the second camera cluster and perform target tracking to generate target positioning data.

It should be noted that the second tracking and positioning system and the second camera cluster can be attached to a monitoring service system outside the service system, and the two service systems only perform limited data exchange during initial modeling and target verification. The external camera cluster does not really exist in the local service system, but is just a virtual mapping camera in the mapping model. By verifying the consistency between the virtual positioning data and the real positioning data, the authenticity and effectiveness of target tracking can be verified, and strict process control of the target tracking task can be realized.

3 is a sequence diagram 2 of a virtual positioning method according to an embodiment of the present application. As shown in FIG. 3 , in addition to the virtual positioning method in FIG. 2 , the virtual positioning method also includes a process of verifying virtual positioning data; Specifically, as shown in FIG. 3 , the second camera cluster will send the collected frame image information stream to the second tracking and positioning system, and the second tracking and positioning system will generate actual positioning data based on the received frame image information stream (ie , target positioning data), and send the actual positioning data to the service system, and the service system can verify the virtual positioning data based on the actual positioning data.

That is, the second tracking and positioning system receives the positioning data of the second camera cluster of the target, and verifies the virtual positioning data of the target and the positioning data of the second camera cluster according to a predetermined rule, and generates a verification of the positioning data of the target in the second camera cluster result.

In an optional embodiment, the virtual positioning method further includes: in response to the relay tracking request signal, and determining the best sampling camera in the second camera cluster according to the virtual positioning data; and setting the attribute of the best sampling camera as The first camera cluster is to obtain the updated first camera cluster; the image information stream collected by the updated first camera cluster is acquired; the image information stream is sent to the positioning system, wherein the positioning system generates a target based on the image information stream after the update The object localization data in the first camera cluster of .

In this embodiment, the best sampling camera in the second camera cluster can be determined according to the virtual positioning data of the second camera cluster in response to the relay tracking request of the target tracking task, and the best camera attribute can be converted into the first camera cluster ; wherein, after the best sampling camera attribute here is converted into the first camera cluster, it is used to replace the original tracking task on-site sampling camera, and relay the frame image information sampling of the target tracking task.

It should be noted that, in this embodiment of the present application, the starting condition of the target tracking task relay tracking request is that the target is located in the predetermined viewing area of the first camera cluster, and/or the target determination accuracy is lower than a threshold. A detailed description will be given below.

In one aspect, in response to the relay tracking request signal, the method includes: determining the distribution density of cameras in the predetermined application scene where the first camera cluster and the second camera cluster are located; when determining that the distribution density is less than a predetermined value, when the target is located in the viewing area of the first camera cluster. When the edge is at a predetermined position, in response to a relay tracking request signal; or, when it is determined that the distribution density is not less than a predetermined value, when the target leaves the middle position of the viewing area of the first camera cluster, in response to a relay tracking request signal.

In this embodiment, the relay tracking request of the target tracking task may set the start standard of the relay tracking request according to the density of the camera site arrangement. For example, in a scene with a low camera density, the overlapping degree of the viewing areas of each camera is low, and a relay tracking request can be initiated when the tracked target is located in a predetermined range at the edge of the sampling area of the first camera cluster sampling camera; for example, in a scene with a high camera density , the framing area of each camera has a high degree of overlap, and it can be set that the relay tracking request is started when the tracked target leaves the predetermined range in the middle of the framing area of the sampling camera. For example, the camera density of the chain image acquisition device is very high, and a relay tracking request condition with a high starting standard can be set to realize the mirror-based target tracking record.

In another aspect, responding to the relay tracking request signal includes: determining that the positioning accuracy of the target is lower than a predetermined threshold; and responding to the relay tracking request signal.

4 is a sequence diagram 3 of a virtual positioning method according to an embodiment of the present application. As shown in FIG. 4 , in addition to the virtual positioning method in FIG. 2 , it also includes the following steps: the service system starts a target tracking task relay tracking request, and then Determine the best sampled camera in the second camera cluster, and send the specified best sampled camera to the second camera cluster, the second camera cluster will convert the attributes of the best sampled camera to the first camera cluster, and receive the updated For the frame image information stream sent by the first camera cluster, the service system will send the received frame image information stream to the tracking and positioning system, and the tracking and positioning system will generate target positioning data according to the updated frame image information stream sent by the first camera cluster. , and send the target positioning data to the service system.

Through the relay tracking in the embodiment of the present application, the continuity of the target tracking task can be ensured and the target positioning accuracy can be improved.

In an optional embodiment, after the virtual positioning of the target is performed according to the virtual positioning data, the virtual positioning method may further include: in response to the identification matching task, determining that the second camera cluster includes at least one camera of the target; Acquiring frame image information of at least one camera; identifying feature identifiers in the frame image information, and generating target identifier information of the target based on the feature identifiers.

In this embodiment, the identification features in the frame image information can be identified as image features, for example, facial features of people, contour features of objects, color features, barcodes or two-dimensional codes, etc.; the feature identification can be a combination of multiple identifications Matching is associated with the same tracked target. For example, when performing feature recognition on a car, the color identification of the car can be associated with the color feature, the type identification of the car can be associated with the contour feature, and the license plate number identification of the associated car can be associated with the license plate recognition.

It should be noted that, in the actual application process, other non-image features, such as RFID radio frequency features, sound features, visible light stroboscopic features or motion features, can also be matched and correlated.

In addition, the service system can set different starting conditions for the identification matching task, for example, the target matching task sets a fixed start interval, for example, in response to the target determination accuracy of the target tracking task being lower than the threshold, for example, in response to the target tracking A target in a task is lost, eg, in response to a user terminal's identity verification command, and so on. When the identification verification task is started, the service system determines the camera including the target in the second camera cluster through the virtual positioning data of the second camera cluster, and the main operation requests the frame image buffer information of the camera including the target. The identification verification task analyzes the frame image information of the cameras containing the target in the first camera cluster and the second camera cluster, generates target identification information, matches the tracked target, improves the accuracy of the tracking result, and reduces the target tracking task bias. Causes the tracking target and the target identification to be incorrectly associated and matched, which is beneficial to the retrieval query of the target.

FIG. 5 is a fourth sequence diagram of a virtual positioning method according to an embodiment of the present application. As shown in FIG. 5 , in addition to the virtual positioning method flow shown in FIG. 2 , the following flow may also be included: the service system may start an identification matching task. , and based on the identification matching task to determine the camera that contains the target in the second camera cluster, then request the frame image information of the specified camera from the second camera cluster, and identify the identification features in the frame image information to generate the identification information of the target.

In an optional embodiment, performing virtual positioning on the target according to the virtual positioning data includes: responding to a target query instruction of the terminal device, and obtaining target identification information of the target query instruction; combining the target positioning data and the target identification information based on the target identification information /or virtual positioning data to determine the viewfinder camera; and feed back the video stream of the target collected by the viewfinder camera to the terminal device.

In this embodiment, the target query instruction of the user terminal can be responded to, and the target identifier of the query instruction can be matched; according to the positioning data of the first camera cluster and/or the virtual positioning data of the second camera cluster, the framing is determined according to predetermined rules lens; the frame image information collected by the viewfinder camera is sent to the user terminal for the user terminal to query.

6 is a sequence diagram 5 of a virtual positioning method according to an embodiment of the present application. As shown in FIG. 6 , in addition to the method shown in FIG. 2 , the virtual positioning method further includes the following steps: the service system generates target identification information, And after receiving the matching queried target identifier of the user terminal, determine the preferred viewfinder camera; obtain the frame image information flow of the viewfinder camera in the determined first camera cluster, and/or, obtain the determined viewfinder camera in the second camera cluster The frame image information stream generated is further sent to the user terminal based on the generated video information stream; wherein, the user terminal will preferably locate the viewfinder camera in the first camera cluster or the second camera cluster.

In an optional embodiment, the virtual positioning method provided in this embodiment of the present application may also respond to an instruction of switching the display angle of the user terminal, and according to the positioning data of the first camera cluster, and/or the second camera cluster For virtual positioning data, the viewfinder camera is switched according to a predetermined rule; the frame image information collected by the switched viewfinder camera is sent to the user terminal.

7 is a sequence diagram 6 of a virtual positioning method according to an embodiment of the present application. As shown in FIG. 7 , the method is basically the same as that in FIG. 6 . Specifically, the service system in FIG.

In an optional embodiment, before performing virtual positioning on the target according to the virtual positioning data, the virtual positioning method may further include: in response to the tracking task request, determining a positioning calibration camera in the second camera cluster according to the virtual positioning data ; Obtain the frame image cache information of the positioning calibration camera; generate the calibration positioning data of the target positioning and calibrating the camera based on the frame image cache information; modify the target positioning data according to the calibration positioning data.

In this embodiment, the target can be assisted in positioning; specifically, in response to the target tracking task request, the positioning calibration camera in the second camera cluster can be determined according to the virtual positioning data of the second camera cluster; Frame image cache information; generate the positioning data of the target positioning calibration camera according to the frame image cache information; correct the target positioning data according to the positioning data of the positioning calibration camera.

Wherein, the starting condition of the tracking task request includes one of the following: the target is lost, the determination accuracy of the target is lower than a predetermined threshold, and the predetermined plan of the target tracking task.

FIG. 8 is a sequence diagram 7 of a virtual positioning method according to an embodiment of the present application. As shown in FIG. 8 , the processing of the virtual positioning method includes the steps shown in FIG. 2 , and may also include the following steps: the service system is based on the target tracking task Start the assisted positioning request, determine the positioning calibration camera in the second camera cluster, request the frame image cache information of the specified camera from the second camera cluster, and obtain the frame image cache information fed back by the second camera cluster. The service system is based on the frame image cache. Information Send the virtual positioning data and frame image buffer information of the positioning and calibration camera to the tracking and positioning system to generate the positioning data of the target in the positioning and calibration camera, and send the positioning data of the target in the positioning and calibration camera to the service system, and the service system is based on the positioning data. Corrected targeting data.

That is, in this embodiment, the service system corrects the target positioning data according to the positioning data of the target in the positioning calibration camera, so as to realize the recovery of target tracking; by assisting the positioning, it can solve the problem that the existing technology is blocked when tracking the target and other factors that affect the tracking When it is lost, it is difficult to continue tracking, and it is easy to lose the original tracking target; the assisted positioning request can also be the request plan task of the target tracking task. Camera frame image information to improve the accuracy of tracking results.

In an optional embodiment, the virtual positioning method may further include: determining a viewfinder camera according to a predetermined rule according to target positioning data and/or virtual positioning data; and sending frame image information collected by the viewfinder camera to a predetermined storage medium.

In this embodiment, a viewfinder camera may be determined according to a predetermined rule according to the positioning data of the first camera cluster and/or the virtual positioning data of the second camera cluster, and frame image information collected by the viewfinder camera is sent to the storage system.

9 is a sequence diagram 8 of a virtual positioning method according to an embodiment of the present application. As shown in FIG. 9 , the processing of the virtual positioning method includes, in addition to the steps shown in FIG. 2 , the following steps: generating a target identifier by using a service system information, and generate the best view camera based on the identification information, and select the acquisition cameras in the first camera cluster and the second camera cluster, and obtain the frame image information flow of the sampling camera, and generate the video information flow based on the obtained frame image information flow. , and send the generated video information stream to the user terminal.

The embodiments of the present application will be described below with reference to different scenarios.

Scenario Example 1

Figure 10(a) is a schematic diagram 1 of a scenic spot according to an embodiment of the present application, the scenic spot provides a kind of MV art short film service for tourists, and the service system of the scenic spot can automatically generate MV art with the tourists as the protagonists after recognizing the facial features of the tourists Short video. Among them, the initial part of the MV art short film template is set with the video images of tourists walking in front of each scenic spot, and the service system searches the video images containing the facial features of the tourists in the video database of the camera collection group of each scenic spot. Insert the retrieved video clips containing facial features into the beginning of the MV art clip. Among them, the end part of the MV art short template is set with the video images of tourists walking on the back of each scenic spot, and the service system cannot retrieve the back images of tourists without facial features through the existing technology.

The following content is the implementation process of the service system matching the virtual identification information of tourists' bodies for video images.

Fig. 10(b) is a schematic diagram 2 of a scenic spot according to an embodiment of the present application. As shown in Fig. 10(b), the corridors of the scenic spot are respectively equipped with three framing cameras camera1, camera2 and camera3, and the mapping model system includes three framing cameras 3D mapping model data for the camera. The three-dimensional mapping model data can have various modeling methods. For example, a triangle template is placed between the three viewfinder cameras. According to the actual shape and size of the triangle template and the three corners of the triangle in the viewfinder image of each camera, in the viewfinder image The corresponding function mapping relationship of the three cameras in the three-dimensional space can be constructed. The process of initializing modeling of the mapping model system is not specifically limited in this embodiment of the present application.

In this embodiment, camera1 and camera2 are set as the first camera cluster, and camera3 is set as the second camera cluster. The tracking and positioning system processes the frame image information collected by camera1 and camera2 in the first camera cluster to generate positioning data for tourists. The service system matches the tourist identification information to the positioning data through facial feature recognition.

In addition, the service system here extracts the three-dimensional model data of the first camera cluster and the second camera cluster in the mapping model system, and obtains the three-dimensional mapping relationship of camera1, camera2 and camera3; and the service system extracts the tourist positioning data generated by the tracking and positioning system, combined with The three-dimensional mapping relationship is calculated to obtain the virtual positioning data corresponding to the tourists in camera3.

In addition, by retrieving the virtual positioning data, the service system can obtain the frame image information including the back pictures of tourists, so as to meet the material requirements of the MV art short film service. It is also possible to set different retrieval conditions for virtual positioning data to meet different material requirements. For example, if the virtual positioning data for retrieval is set to match the positioning of the tourist’s face in the middle of the viewfinder, the image material showing the upper part of the body can be obtained; set the retrieval virtual positioning If the data is consistent with the tourist's face being positioned in the predetermined interval above the outside of the viewfinder, the image material that only shows the feet and legs can be obtained.

Scenario Example 2

11 is a schematic diagram of a logistics transfer station according to an embodiment of the present application. As shown in the figure, a material monitoring and handover service system is deployed in the logistics transfer station in the goods receiving and dispatching link, which can complete the tracking and monitoring system for materials in the transfer station and the transport vehicle monitoring system. Seamless handover can realize the full monitoring and traceability of materials in the logistics process.

In the embodiment shown in the figure, the logistics transfer station and the transport vehicle are each deployed with a full-process monitoring system, and the first tracking and positioning system in the two sets of monitoring systems is deployed at the logistics transfer station, which is the local tracking and positioning system of the material monitoring and handover service system; The second tracking and positioning system in the two sets of monitoring systems is deployed on the transport vehicle and is the flow tracking and positioning system of the material monitoring and handover service system. In the embodiment, the camera of the first camera cluster is installed at the position of the material conveying device of the logistics transfer station, and is configured as the sampling camera of the first tracking and positioning system, and the first tracking and positioning system is configured to generate the positioning data of the first camera cluster of the target; transportation; A camera of the second camera cluster is installed in the cargo compartment of the vehicle, and is configured as a sampling camera of the second tracking and positioning system, and the second tracking and positioning system is configured to generate positioning data of the second camera cluster of the target. The second camera cluster is an external camera of the local monitoring system of the transfer station, and its attribute in the local monitoring system of the transfer station is a local virtual camera, not a local entity camera; the attribute of the second camera cluster in the monitoring system of the transport vehicle is the first The camera cluster is the local physical camera of the monitoring system of the transport vehicle.

It should be noted that the first tracking and positioning system and the second tracking and positioning system match the identification information by scanning the barcode of the material. The barcode of the material is generated by the logistics worker manually scanning the material label with the barcode scanner at the handover position of the material. .

The local mapping model system needs to perform initial modeling before monitoring material handover, and establish two-dimensional model data of the first camera cluster and the second camera cluster. The two-dimensional mapping data is that the mapping model constructs the mapping relationship corresponding to each camera in the two-dimensional model by performing overlapping analysis on the image information collected by the first camera cluster and the second camera cluster. The overlapping analysis of the image information in this embodiment is to analyze the partial overlapping area in the image information. As shown in the figure, most of the viewing area of the two viewfinder cameras is covered by obstructions such as the cargo door, and the two sets of images Only a small area of the information overlaps. During the initial modeling process of the mapping model system of this embodiment, on the one hand, the influence of occluders on the modeling process should be avoided, and on the other hand, the influence of different brightness of the camera viewing environment inside and outside the vehicle on the modeling process should be avoided.

The material monitoring and handover service system also has the following functions: extracting the two-dimensional model data in the mapping model system, obtaining the two-dimensional mapping relationship between the first camera cluster and the second camera cluster; extracting the material positioning data generated by the first tracking and positioning system , Combined with the two-dimensional mapping relationship, the virtual positioning data corresponding to the material in the second camera cluster is obtained through calculation processing; the positioning data of the material in the second camera cluster generated by the second tracking and positioning system deployed on the transport vehicle is received, and the calibration Check the consistency between the positioning data of the material in the second camera cluster and the virtual positioning data. If the consistency is greater than the predetermined threshold, it is determined that the process of material handover in the two monitoring systems is complete, and the material monitoring and handover system is determined. The local tracking and positioning system and the flow tracking and positioning system are verified by material handover. Two sets of tracking and positioning systems mutually verify the tracking thread of the material, which is convenient for the synchronous retrieval of material traceability.

Scenario Example 3

12 is a schematic diagram of a traffic monitoring area according to an embodiment of the present application. As shown in the figure, the urban traffic department installs intensive cameras in key traffic monitoring areas to implement tracking monitoring of road vehicles.

In the embodiment, the cameras are dense, and each vehicle is simultaneously in the viewing area of multiple cameras. After the service system starts the tracking task thread for a certain vehicle, the service system samples the tracking task image from one of the cameras of the vehicle in the viewing area according to a predetermined rule, and configures the attribute of the sampling camera as the first camera cluster; The remaining cameras are configured as the second camera cluster of the tracking task thread.

The service system analyzes the characteristics of the tracked target based on the collected images, and can generate three types of target identifications: vehicle color, vehicle type and vehicle number plate. For example, the service system can match the three target identifiers of "NO.002 red car", "NO.005 car" and "88888 license plate" to the target at the same time by identifying the features of the sampled images. The service system can use these three identification information Retrieve the monitoring records of the vehicle.

The mapping model system in the service system generates the two-dimensional model data of all cameras by initializing the modeling. The corresponding mapping relationship in the 2D model.

According to the positioning data of the tracked vehicle in the first camera cluster generated by the tracking and positioning system, combined with the two-dimensional model data in the extraction and mapping model system, the virtual positioning data corresponding to the tracked vehicle in the second camera cluster is obtained through computational processing. . Due to the dense number of cameras, the target of the tracked vehicle is matched with the virtual positioning data of multiple cameras at the same time.

When the tracked vehicle travels to the predetermined range of the sampling camera viewing area, continuing to use the camera as the sampling camera of the tracking task thread will affect the accuracy of target tracking and positioning, and may also lead to the loss of the tracked target. In response to the relay tracking request of the target tracking task, the service system determines the best sampling camera in the second camera cluster according to the virtual positioning data, and converts the camera attributes into the first camera cluster, which is used to replace the original tracking task thread Sampling the camera to sample the frame image information of the relay target tracking task. Through relay tracking, the continuity of target tracking tasks can be ensured.

In this embodiment, the identification matching task of the service system is started every 5 seconds, which reduces the deviation of the target tracking task and causes the tracking target and the target identification to be erroneously associated and matched. When the identification verification task is started, the service system determines the camera including the target in the second camera cluster through the virtual positioning data of the second camera cluster, and the main operator requests the frame image information of the camera including the target. The identification verification task analyzes the frame image information of the cameras containing the target in the first camera cluster and the second camera cluster, generates target identification information, matches the tracked target, and improves the accuracy of the tracking result.

Scenario Example 4

13 is a schematic diagram of a kindergarten according to an embodiment of the present application. As shown in the figure, the kindergarten provides parents with a real-time video service for viewing children's activities from multiple perspectives. All public places of the kindergarten are equipped with chain image acquisition devices.

The distance between each camera in the chain image acquisition device is 0.2 meters, and the mapping model system can construct the corresponding mapping relationship of each camera in the two-dimensional model by analyzing the overlapping area through the image information collected by the existing cameras.

In the embodiment, the service system determines 1 camera as the target tracking task sampling camera among the 20 adjacent cameras in the chain image acquisition device, and the attribute of the target tracking task sampling camera is configured as the first camera cluster, and is not configured as the first camera cluster. The other cameras of the first camera cluster are configured as the second camera cluster. In an application scenario where cameras are densely arranged on-site, if each camera is used for sampling for target tracking tasks, on the one hand, the system is unaffordable, and on the other hand, over-sampling for tracking and positioning tasks results in a serious waste of system resources.

In the embodiment, the tracking and positioning system includes multiple sets of target detection algorithms, including frame difference method, background subtraction method, expectation maximization method, optical flow method, statistical model method, level set method, etc., which can be switched in real time according to scene requirements. Object detection algorithm. The service system pre-stores the facial features of the tracked object, performs identification verification on the tracked positioning target according to a predetermined calculation frequency, and realizes the identity identification matching of the tracked positioning target.

In the embodiment, after receiving the positioning data of each identity mark, the virtual positioning system extracts the two-dimensional model data of the first camera cluster and the second camera cluster in the mapping model system, obtains the two-dimensional mapping relationship of all the cameras, and through calculation processing, The corresponding virtual positioning data of each camera in the second camera cluster of each identity identifier is obtained.

As shown in the figure, the service system receives the target query instruction generated by the child's parent through the operation of the user terminal, and in response to the target query instruction, matches the identity identifier of the query instruction (that is, matches the facial feature identifier that the parent needs to query the child's target). By retrieving the positioning data and virtual positioning data corresponding to the identity identifier, the service system determines the camera that contains the target within the viewing range (that is, determines that the viewing range includes the camera of the child to be queried). The preferred camera with the target closest to the middle area of the framing range is selected from the camera cluster including the target within the framing range as the framing camera. The frame image information is extracted from the determined preferred viewing camera, converted into a video information stream and sent to the user terminal for display to the user.

As shown in the figure, the service system receives an instruction to switch the display angle of view generated by the parent of the child through the user terminal operation. The switching direction of the viewing angle command switches the camera that includes the target within the viewing range to be the viewing camera; the frame image collected by the viewing camera is converted into a video information stream and sent to the user terminal for display to the operator.

Scenario Example 5

14 is a schematic diagram of a subway operation company according to an embodiment of the present application. As shown in the figure, the subway operation company deploys a multi-view mirror-type monitoring and recording system. Multi-angle mirror monitoring record of the tracked target. The chain image acquisition device collects frame image information through different cameras, and then samples the frame image information clusters collected by different cameras through predetermined rules to generate video, which can realize the video monitoring record of the lens moving with the target. By deploying the chain image acquisition device, the subway operating company can start an uninterrupted mirror-based monitoring and recording thread for each passenger entering the subway station. One subway station → take the subway train → get off the subway train and enter the second subway station → leave the entire whereabouts process of the second subway station. Each mirror-based monitoring and recording thread corresponds to a tracked and recorded target, and multiple mirror-based monitoring and recording threads correspond to multiple tracked and recorded targets respectively. The number of start-up recording threads is determined by the computing power of the service system. Intensive deployment of multiple chain image acquisition devices can realize multi-angle tracking video recording of passengers entering the subway station. A mirror-type monitoring and recording thread can contain images from multiple angles, and corresponding system managers can view it in real time at the same time. And query to tracked targets from multiple angles of the mirror monitoring records.

Since the subway station is a public place, according to the relevant privacy protection laws in some areas, the face recognition technology cannot be used for the public in the public place. The identification verification device in this embodiment is set at the gate of the subway station, and the identification verification device is The RFID identification of the subway ticket card or the image code identification of the ticket card. Querying the tracking records of a certain target through the service system can only be retrieved by the corresponding identification of the ticket card, and cannot use the target features involving personal privacy, including face recognition, as the identification for retrieval.

The distance between each camera in the chain image acquisition device is 0.4 meters, and a 1fps low frame rate lens dynamic tracking video can be performed in the viewing area. The framing lens moves 0.4 meters per second with a mirror-like video presentation effect.

The mapping model system analyzes the image information collected at the same time by all the cameras of each chain image acquisition device through the artificial intelligence system, and constructs the corresponding mapping relationship of all the cameras of all chain image acquisition devices in the system in the three-dimensional model. Dynamic models are also preset in the mapping model system, and the dynamic models are different three-dimensional models corresponding to different conditions. The mapping relationship between the camera clusters in the subway station and the camera clusters in different subway trains, as well as the mapping relationship between the camera clusters in the subway train and different subway station camera clusters, are selected after the preconditions for the subway routine car to stop at the subway station are determined. .

In the embodiment, the chain image acquisition devices are densely distributed, and the chain image acquisition devices in opposite viewing directions are relatively parallel distributed. The service system selects the sampling camera for the target tracking task through artificial intelligence analysis, and the attributes of the target tracking task sampling camera are configured as: The first camera cluster, and other cameras not configured as the first camera cluster are configured as the second camera cluster. By selecting the sampling camera to determine the target task through artificial intelligence analysis, the target can be tracked and sampled with the least number of cameras, reducing the calculation amount of the system to perform the target tracking task and reducing the system load. The tracking and positioning system uses the deep neural network to track and locate each target in each place, and generates the positioning data of each target.

In the embodiment, the virtual positioning system extracts the dynamic model data of the first camera cluster and the second camera cluster in the mapping model system, and after the subway train stops at the platform, updates the three-dimensional mapping relationship of all cameras under the currently determined conditions. The virtual positioning system obtains the positioning data of each target generated by the tracking and positioning system, and combines the three-dimensional mapping relationship to obtain the corresponding virtual positioning data of each target in each camera in the second camera cluster through calculation processing.

According to the positioning data of the first camera cluster and the virtual positioning data of the second camera cluster, corresponding to the target tracking task, the service system determines the tracked target and the camera that contains the target within the viewing range, and selects the best camera for each image acquisition chain. Viewfinder camera.

The service system obtains the frame image information of the best view camera, converts it into a video information stream and sends it to the storage system. The storage system stores the received video information stream for serving the retrieval query of system administrators.

In the target tracking task, if the tracking target is lost or the target determination accuracy is lower than the threshold, the target tracking task starts to initiate an assisting positioning request. In response to the auxiliary positioning request, the service system determines, according to the virtual positioning data, a camera in the second camera cluster that meets the predetermined condition as a positioning calibration camera, obtains the frame image cache information of the positioning calibration camera, and obtains the virtual positioning data and frame image of the positioning calibration camera according to the positioning calibration camera. The cached information is used for target tracking and positioning analysis, and the positioning data of the target in the positioning calibration camera is generated. The service system corrects the target positioning data according to the positioning data of the target in the positioning calibration camera, and realizes the recovery of target tracking. By assisting the positioning, it can solve the problem that it is difficult to continue the tracking when the tracking is lost due to factors such as being occluded when the tracking target is being tracked, and the original tracking target is easily lost. The assisted positioning request may also be a request plan task of the target tracking task, and the target tracking task is configured with a plan for regularly starting the target tracking auxiliary request, and dynamically extracts image information of different camera frames to improve the accuracy of the tracking result.

According to another aspect of the embodiment of the present application, a virtual positioning device is also provided. FIG. 15 is a schematic diagram of the virtual positioning device according to the embodiment of the present application. As shown in FIG. 15 , the virtual positioning device includes: an obtaining unit 1501, The sending unit 1503 and the virtual positioning unit 1505. The virtual positioning device will be described below.

The obtaining unit 1501 is configured to obtain target positioning data of a target in a first camera cluster, wherein the first camera cluster is a camera used in a positioning system, and the positioning system is used to perform a target tracking task to generate positioning data.

The sending unit 1503 is configured to send the target positioning data to the positioning system, wherein the positioning system uses the target mapping model and the target positioning data to generate virtual positioning data of the target in the second camera cluster, and the target mapping model is used to describe the first camera cluster The mapping relationship with the second camera cluster in space.

The virtual positioning unit 1505 is configured to perform virtual positioning on the target according to the virtual positioning data.

It should be noted here that the above obtaining unit 1501, sending unit 1503 and virtual positioning unit 1505 correspond to steps S102 to S106 in Embodiment 1, and the examples and application scenarios implemented by the above units and corresponding steps are the same, but not limited to The content disclosed in Example 1 above. It should be noted that the above-mentioned units may be executed in a computer system such as a set of computer-executable instructions as part of an apparatus.

As can be seen from the above, in the above-mentioned embodiments of the present application, the target positioning data of the target in the first camera cluster can be obtained by using the acquisition unit, wherein the first camera cluster is the camera used in the positioning system, and the positioning system is used for Perform the target tracking task to generate positioning data; then use the sending unit to send the target positioning data to the positioning system, wherein the positioning system uses the target mapping model and the target positioning data to generate virtual positioning data of the target in the second camera cluster, and the target mapping model It is used to describe the mapping relationship between the first camera cluster and the second camera cluster in space; and the virtual positioning unit is used to perform virtual positioning on the target according to the virtual positioning data. The virtual positioning device provided by the embodiment of the present application achieves the purpose of locating the target through the camera cluster, achieves the technical effect of improving the accuracy of locating the target, and solves the problem that the application of the image sensor in the related art is relatively simple and cannot be Technical issues of large-scale collaborative tracking using image sensors.

According to another aspect of the embodiments of the present application, a virtual positioning device is also provided, including: an obtaining unit configured to obtain target positioning data of a target in a first camera cluster, wherein the first camera cluster is used in the positioning system For the camera used, the positioning system is used to perform the target tracking task to generate positioning data; the sending unit is configured to send the target positioning data to the positioning system, wherein the positioning system uses the target mapping model and the target positioning data to generate the target in the second. The virtual positioning data in the camera cluster, the target mapping model is used to describe the spatial mapping relationship between the first camera cluster and the second camera cluster; the virtual positioning unit is configured to perform virtual positioning on the target according to the virtual positioning data.

In an optional embodiment, the virtual positioning device further includes: a verification unit, configured to verify the virtual positioning data before performing virtual positioning on the target according to the virtual positioning data; wherein the verification unit includes : a first acquisition module, configured to acquire the actual positioning data of the target in the second camera cluster; a verification module, configured to use a predetermined verification rule to determine the similarity between the actual positioning data and the virtual positioning data, so as to compare the virtual positioning data with the virtual positioning data. The consistency before the actual positioning data is checked.

In an optional embodiment, the virtual positioning device further includes: a first determining unit, configured to respond to the relay tracking request signal and determine the best sampling camera in the second camera cluster according to the virtual positioning data; a setting unit , set to set the attribute of the best sampling camera as the first camera cluster, and obtain the updated first camera cluster; the acquisition unit, set to obtain the image information stream collected by the updated first camera cluster; the sending unit, set to The image information stream is sent to the positioning system, wherein the positioning system generates target positioning data of the target in the updated first camera cluster based on the image information stream.

In an optional embodiment, the first determination unit includes: a first determination module configured to determine the distribution density of cameras in the predetermined application scenarios where the first camera cluster and the second camera cluster are located; the second determination module configured to determine In order to determine that when the distribution density is less than a predetermined value, when the target is located at a predetermined position on the edge of the viewing area of the first camera cluster, in response to the relay tracking request signal; or, the third determining module is set to determine that the distribution density is not less than the predetermined value. When the target leaves the middle position of the viewing area of the first camera cluster, it responds to the relay tracking request signal.

In an optional embodiment, the first determination unit includes: a fourth determination module configured to determine that the positioning accuracy of the target is lower than a predetermined threshold; and a response module configured to respond to a relay tracking request signal.

In an optional embodiment, the virtual positioning device further includes: a second determining unit, configured to, after virtual positioning of the target according to the virtual positioning data, in response to the identification matching task, determine that the second camera cluster contains at least one camera of the target; an acquiring unit configured to acquire frame image information of the at least one camera; and a first generating unit configured to recognize feature identifiers in the frame image information, and generate target identifier information of the target based on the feature identifiers.

In an optional embodiment, the virtual positioning unit includes: a second obtaining module, configured to respond to a target query instruction of the terminal device, and obtain target identification information of the target query command; a fifth determining module, configured to be based on The target identification information is combined with target positioning data and/or virtual positioning data to determine the viewfinder camera; the feedback module is configured to feed back the video stream of the target collected by the viewfinder camera to the terminal device.

In an optional embodiment, the virtual positioning device further includes: a third determining unit, configured to determine the second camera according to the virtual positioning data in response to the tracking task request before performing virtual positioning on the target according to the virtual positioning data a positioning calibration camera in the cluster; an acquiring unit, set to obtain frame image cache information of the positioning calibration camera; a second generating unit, set to generate calibration positioning data of the target locating and calibrating the camera based on the frame image cache information; a correction unit, set to Correct the target positioning data according to the calibration positioning data.

In an optional embodiment, the starting condition of the tracking task request includes one of the following: the target is lost, the determination accuracy of the target is lower than a predetermined threshold, and the predetermined plan of the target tracking task.

In an optional embodiment, the virtual positioning device further includes: a fourth determining unit, configured to determine a viewfinder camera according to a predetermined rule according to the target positioning data and/or the virtual positioning data; a sending unit configured to determine the viewfinder camera according to a predetermined rule; The captured frame image information is sent to a predetermined storage medium.

According to another aspect of the embodiments of the present application, a computer-readable storage medium is also provided, where the computer-readable storage medium includes a stored computer program, wherein when the computer program is run by the processor, the device where the computer storage medium is located is controlled to execute the above-mentioned The virtual positioning method of any one.

According to another aspect of the embodiments of the present application, a processor is also provided, where the processor is configured to run a computer program, wherein when the computer program runs, any one of the above virtual positioning methods is executed.

According to another aspect of the embodiments of the present application, a virtual positioning system is also provided, and the virtual positioning system uses any one of the above virtual positioning methods.

The above-mentioned serial numbers of the embodiments of the present application are only for description, and do not represent the advantages or disadvantages of the embodiments.

In the above-mentioned embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.

In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are only illustrative, for example, the division of the units may be a logical function division, and there may be other division methods in actual implementation, for example, multiple units or components may be combined or Integration into another system, or some features can be ignored, or not implemented. On the other hand, the shown or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection of units or modules, and may be in electrical or other forms.

The units described as separate components may or may not be physically separated, and components shown as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.

In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. The above-mentioned integrated units may be implemented in the form of hardware, or may be implemented in the form of software functional units.

The integrated unit, if implemented in the form of a software functional unit and sold or used as an independent product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the present application can be embodied in the form of software products in essence, or the parts that contribute to the prior art, or all or part of the technical solutions, and the computer software products are stored in a storage medium , including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes .

The above are only the preferred embodiments of the present application. It should be pointed out that for those skilled in the art, without departing from the principles of the present application, several improvements and modifications can also be made. It should be regarded as the protection scope of this application.

Claims

A virtual positioning method, comprising:

obtaining target positioning data of the target in a first camera cluster, wherein the first camera cluster is a camera used in a positioning system used to perform a target tracking task to generate positioning data;

The target positioning data is sent to the positioning system, wherein the positioning system uses the target mapping model and the target positioning data to generate virtual positioning data for the target in the second camera cluster, and the target mapping model uses for describing the spatial mapping relationship between the first camera cluster and the second camera cluster;

Perform virtual positioning on the target according to the virtual positioning data.
The method according to claim 1, wherein the target mapping model is a mapping model obtained by the positioning system from a mapping model system, wherein the mapping model system generates a mapping model that includes a predetermined application scenario by means of initialization modeling. The mapping model of the relationship between each camera position and viewing angle.
The method according to claim 1, wherein before performing virtual positioning on the target according to the virtual positioning data, the method further comprises: verifying the virtual positioning data;

Wherein, verifying the virtual positioning data includes:

obtaining the actual positioning data of the target in the second camera cluster;

The similarity between the actual positioning data and the virtual positioning data is determined by using a predetermined check rule, so as to check the previous consistency between the virtual positioning data and the actual positioning data.
The method of claim 1, further comprising:

determining the best sampling camera in the second camera cluster according to the virtual positioning data in response to the relay tracking request signal;

Setting the attribute of the best sampling camera as the first camera cluster to obtain the updated first camera cluster;

acquiring the image information stream collected by the updated first camera cluster;

Sending the image information stream to the positioning system, wherein the positioning system generates target positioning data of the target in the updated first camera cluster based on the image information stream.
The method of claim 4, wherein in response to the relay tracking request signal, comprising:

determining the distribution density of cameras in the predetermined application scenario where the first camera cluster and the second camera cluster are located;

When it is determined that the distribution density is less than a predetermined value, when the target is located at a predetermined position on the edge of the viewing area of the first camera cluster, in response to the relay tracking request signal; or,

When it is determined that the distribution density is not less than a predetermined value, when the target leaves the middle position of the viewing area of the first camera cluster, the relay tracking request signal is responded to.
The method of claim 5, wherein in response to the relay tracking request signal, comprising:

determining that the positioning accuracy of the target is lower than a predetermined threshold;

in response to the relay tracking request signal.
The method according to claim 1, wherein after performing virtual positioning on the target according to the virtual positioning data, the method further comprises:

In response to the identification matching task, determining that the second camera cluster includes at least one camera of the target;

acquiring frame image information of the at least one camera;

Identify the feature identification in the frame image information, and generate target identification information of the target based on the feature identification.
The method according to claim 7, wherein the virtual positioning of the target according to the virtual positioning data comprises:

In response to the target query instruction of the terminal device, and obtain the target identification information of the target query instruction;

determining a viewfinder camera based on the target identification information in combination with the target positioning data and/or the virtual positioning data;

The video stream of the target captured by the viewfinder camera is fed back to the terminal device.
The method according to any one of claims 1 to 8, wherein before performing virtual positioning on the target according to the virtual positioning data, the method further comprises:

In response to the tracking task request, determining a positioning calibration camera in the second camera cluster according to the virtual positioning data;

acquiring the frame image buffer information of the positioning calibration camera;

generating calibration positioning data of the target in the positioning calibration camera based on the frame image buffer information;

The target positioning data is corrected according to the calibration positioning data.
The method according to claim 9, wherein the starting condition of the tracking task request comprises one of the following: the target is lost, the determination accuracy of the target is lower than a predetermined threshold, and the predetermined plan of the target tracking task.
The method of claim 9, further comprising:

According to the target positioning data and/or the virtual positioning data, the viewfinder camera is determined according to a predetermined rule;

The frame image information collected by the viewfinder camera is sent to a predetermined storage medium.
A virtual positioning device, comprising:

an acquisition unit configured to acquire target positioning data of a target in a first camera cluster, wherein the first camera cluster is a camera used in a positioning system, and the positioning system is used to perform a target tracking task to generate positioning data;

A sending unit, configured to send the target positioning data to the positioning system, wherein the positioning system uses the target mapping model and the target positioning data to generate virtual positioning data of the target in the second camera cluster, so The target mapping model is used to describe the mapping relationship between the first camera cluster and the second camera cluster in space;

A virtual positioning unit, configured to perform virtual positioning on the target according to the virtual positioning data.
The device according to claim 12, wherein the target mapping model is a mapping model obtained by the positioning system from a mapping model system, wherein the mapping model system generates a mapping model that includes a predetermined application scenario by means of initialization modeling The mapping model of the relationship between each camera position and viewing angle.
The apparatus according to claim 12, further comprising: a verification unit, configured to verify the virtual positioning data before performing virtual positioning on the target according to the virtual positioning data;

Wherein, the verification unit includes:

a first acquisition module, configured to acquire actual positioning data of the target in the second camera cluster;

A verification module configured to determine the similarity between the actual positioning data and the virtual positioning data by using a predetermined verification rule, so as to verify the previous consistency between the virtual positioning data and the actual positioning data.
The apparatus of claim 12, further comprising:

a first determining unit, configured to respond to a relay tracking request signal and determine the best sampling camera in the second camera cluster according to the virtual positioning data;

a setting unit, configured to set the attribute of the best sampling camera as the first camera cluster, to obtain the updated first camera cluster;

the obtaining unit, configured to obtain the image information stream collected by the updated first camera cluster;

The sending unit is configured to send the image information flow to the positioning system, wherein the positioning system generates the target positioning of the target in the updated first camera cluster based on the image information flow data.
The apparatus according to claim 15, wherein the first determining unit comprises:

a first determining module, configured to determine the distribution density of cameras in a predetermined application scenario where the first camera cluster and the second camera cluster are located;

The second determination module is configured to respond to the relay tracking request signal when it is determined that the distribution density is less than a predetermined value when the target is located at a predetermined position on the edge of the viewing area of the first camera cluster; or,

The third determining module is configured to respond to the relay tracking request signal when the target leaves the middle position of the viewing area of the first camera cluster when it is determined that the distribution density is not less than a predetermined value.
The apparatus according to claim 16, wherein the first determining unit comprises:

a fourth determining module, configured to determine that the positioning accuracy of the target is lower than a predetermined threshold;

A response module configured to respond to the relay tracking request signal.
The apparatus of claim 12, further comprising:

a second determining unit, configured to determine that at least one camera of the target is included in the second camera cluster in response to an identification matching task after performing virtual positioning on the target according to the virtual positioning data;

The acquisition unit is configured to acquire frame image information of the at least one camera;

The first generating unit is configured to recognize a feature identifier in the frame image information, and generate target identifier information of the target based on the feature identifier.
The apparatus according to claim 18, wherein the virtual positioning unit comprises:

A second acquiring module, configured to respond to a target query instruction of the terminal device, and acquire target identification information of the target query instruction;

a fifth determination module, configured to determine a viewfinder camera based on the target identification information in combination with the target positioning data and/or the virtual positioning data;

The feedback module is configured to feed back the video stream of the target collected by the viewfinder camera to the terminal device.
The apparatus of any one of claims 12 to 19, further comprising:

a third determining unit, configured to, in response to a tracking task request, determine a positioning calibration camera in the second camera cluster according to the virtual positioning data before performing virtual positioning on the target according to the virtual positioning data;

The obtaining unit is configured to obtain the frame image buffer information of the positioning calibration camera;

a second generating unit, configured to generate calibration positioning data of the target in the positioning calibration camera based on the frame image buffer information;

A correction unit, configured to correct the target positioning data according to the calibration positioning data.
The apparatus according to claim 20, wherein the starting condition of the tracking task request comprises one of the following: the target is lost, the determination accuracy of the target is lower than a predetermined threshold, and the predetermined plan of the target tracking task.
The apparatus of claim 20, further comprising:

a fourth determining unit, configured to determine a viewfinder camera according to a predetermined rule according to the target positioning data and/or the virtual positioning data;

The sending unit is configured to send the frame image information collected by the viewfinder camera to a predetermined storage medium.
A virtual positioning system using the virtual positioning method described in any one of the above claims 1 to 11.
A computer-readable storage medium, the computer-readable storage medium comprising a stored computer program, wherein when the computer program is run by a processor, a device where the computer storage medium is located is controlled to execute any one of claims 1 to 11 The virtual positioning method described in item.
A processor for running a computer program, wherein the computer program executes the virtual positioning method according to any one of claims 1 to 11 when the computer program runs.