A Local Image Recognition Method and System Based on AR Smart Glasses

Abstract

The present invention discloses a local image recognition method based on an AR smart glasses. The method includes: obtaining dynamic image information within the field of view of the device in real time, analyzing the motion characteristics of the image information to generate a moving object marker; performing high-definition magnified sampling on the local image according to the moving object marker to generate feature information of the moving object marker; performing cross-comparative analysis on the feature information of the moving object marker and preset object information to generate a tracking instruction; sending the tracking instruction to a collaborative device through an intranet and tracking the moving object; realizing a multi-device collaborative object automatic recognition and tracking system based on AR glasses, greatly improving the recognition and tracking efficiency of relevant personnel for a tracking object in occasions with a large number of people in the public, and solving many problems of existing tracking methods.

Description

Technical Field

[0001] The present invention relates to the field of intelligent AR devices, and specifically to a local image recognition method and system based on AR smart glasses. Background Technique

[0002] AR (Augmented Reality, that is, augmented reality) is a technology that combines the real world with virtual information, that is, a technology that enhances the real world with virtual information. It complements and corrects the two kinds of information of the virtual information and the real world, and with the cooperation of various technical means such as multimedia, 3D models, intelligent interaction, and sensors, it achieves the purpose of better supporting work and entertainment.

[0003] In recent years, with the continuous development of AR technology, more and more AR-related technologies and devices have gradually emerged. The functions and user groups of more AR-related devices are more inclined to the civilian consumption level. Some AR devices have certain creator support functions, which can increase the creative behavior methods of creators and stimulate the creative potential and desire of creators. According to the AR devices and technologies in the prior art, it is not difficult for us to find that AR devices are gradually replacing mobile devices in the prior art as the next-generation intelligent terminal, but the potential other application spaces of the still immature AR technology at this stage are still a very broad development opportunity.

[0004] Combining the existing AR smart eye technology with image recognition can be used for the tracking and positioning of moving units, and can effectively assist the work of relevant institutions. In the prior art, when tracking moving units, especially people, most of them are carried out manually by tracking personnel. It is difficult to effectively identify and track the personnel to be tracked in a crowd at a long distance, and the auxiliary tracking monitoring system also has many dead corners and is insensitive to use. Summary of the Invention

[0005] The purpose of the present invention is to provide a local image recognition method and system based on AR smart glasses to solve the problems raised in the above background technique.

[0006] To achieve the above purpose, the present invention provides the following technical solutions:

[0007] A local image recognition method based on AR smart glasses includes the following steps:

[0008] Real-time obtain dynamic image information within the field of view of the device, and perform three-dimensional motion feature analysis on the image information to generate a moving object marker;

[0009] Perform dynamic image acquisition on the moving object marker and extract the feature information of the moving object marker;

[0010] Cross - compare and analyze the characteristic information of the moving object marker with the preset object information to generate a tracking instruction;

[0011] Send the tracking instruction to the collaborative device through the intranet and track the moving object.

[0012] As a further solution of the present invention: The dynamic image information is composed of several consecutive static images collected at a preset time interval, and the preset time interval is used to characterize the movement angular velocity of the device's field of view, and the preset time interval is set in an inverse relationship with the movement angular velocity of the device's field of view.

[0013] As a still further solution of the present invention: The dynamic image information further includes distance data and device position information. The distance data is used to characterize the straight - line distance from each image point on the dynamic image data to the device; The step of performing three - dimensional motion feature analysis on the image information to generate a moving object marker specifically includes:

[0014] Record and store the device position information in real - time, establish a three - dimensional space model, and generate the device movement trajectory;

[0015] Read the static image information in the dynamic image information frame by frame, and obtain the distance information and device position information in the static image information;

[0016] Establish a real - time object model in the three - dimensional space model according to the distance information in the static image information corresponding to the device position information and store it;

[0017] Generate the object model movement trajectory according to the continuously updated real - time object model;

[0018] Mark the moving object model in motion according to the object model movement trajectory.

[0019] As a still further solution of the present invention: The step of performing dynamic image acquisition on the moving object marker and extracting the characteristic information of the moving object marker specifically includes:

[0020] Obtain the relative position and distance data of the moving object marker according to the real - time object model;

[0021] Perform local image high - definition real - time sampling according to the relative position and distance data of the moving object marker to obtain a local dynamic high - definition image;

[0022] Extract features from the local dynamic high - definition image to generate the characteristic information of the moving marker body.

[0023] As a further aspect of the present invention: The method presets preset object information, and the preset object information is the feature information of the tracking object collected in advance. The step of cross-comparing and analyzing the feature information marked on the moving body with the preset object information to generate a tracking instruction specifically includes:

[0024] Establish a feature comparison model according to the preset object information, and the feature comparison model is used to represent the degree of coincidence with the preset object information;

[0025] Obtain the feature information marked on the moving body, and perform comparison and analysis on the feature information marked on the moving body according to the feature comparison model to generate a comparison result. If

[0026] the comparison result is a high degree of coincidence, generate a prompt message, and output the tracking instruction as tracking.

[0027] As a further aspect of the present invention: The step of sending the tracking instruction to the collaborative device through the intranet and tracking the moving body specifically includes:

[0028] When the tracking instruction is tracking, extract the corresponding moving body mark and the tracking instruction;

[0029] Send the moving body mark and the tracking instruction to the collaborative device;

[0030] Continuously track in the three-dimensional space model according to the corresponding moving mark body information.

[0031] As a further aspect of the present invention: The method further includes:

[0032] Receive the moving body mark and the tracking instruction from the collaborative device;

[0033] Retrieve and identify the moving body mark in the three-dimensional space model according to the received moving body mark; if

[0034] the moving body mark in the three-dimensional space model is consistent with the received moving body mark, continuously track in the three-dimensional space model.

[0035] In a second aspect, an embodiment of the present invention aims to provide a local image recognition system based on an AR smart glasses. The local image recognition system specifically includes:

[0036] A global construction module, which is used to obtain dynamic image information within the field of view of the device in real time, and perform motion feature analysis on the image information to generate a moving body mark;

[0037] A sampling recognition module, which is used to perform high-definition magnification sampling on the local image according to the moving body mark to generate the feature information of the moving body mark;

[0038] A tracking and recognition module, which is used to perform cross-comparative analysis on the feature information marked by the moving object and the preset object information to generate a tracking instruction;

[0039] A collaborative tracking module, which is used to send the tracking instruction to the collaborative device through the intranet and track the moving object.

[0040] As a further solution of the present invention: The global construction module specifically includes:

[0041] An environmental image acquisition unit, which is used to obtain dynamic image information within the field of view of the device in real time

[0042] A spatial motion model unit, which is used to record and store the device position information in real time, establish a three-dimensional space model, and generate a device motion trajectory; read the static image information in the dynamic image information frame by frame, and obtain the distance information and device position information in the static image information; establish a real-time object model in the three-dimensional space model according to the distance information in the static image information corresponding to the device position information and store it; generate an object model motion trajectory according to the continuously updated real-time object model;

[0043] A moving object recognition unit, which is used to mark the moving object according to the object model motion trajectory.

[0044] As a further solution of the present invention: The sampling and recognition module specifically includes:

[0045] An individual relative positioning module, which is used to obtain the relative position and distance data of the moving object mark according to the real-time object model;

[0046] An individual image acquisition unit, which is used to perform local image high-definition real-time sampling according to the relative position and distance data of the moving object mark to obtain a local dynamic high-definition image;

[0047] An individual feature acquisition unit, which is used to extract features from the local dynamic high-definition image to generate the feature information of the moving marked body.

[0048] Compared with the prior art, the beneficial effects of the present invention are:

[0049] The present invention provides a local image recognition method and system based on an AR smart glasses, realizing a multi-device collaborative object automatic recognition and tracking system based on AR glasses, greatly improving the recognition and tracking efficiency of relevant personnel for the person to be tracked in occasions with a large number of people in the public, and solving many problems of the existing tracking methods. Description of the Drawings

[0050] Figure 1 It is a flow chart of a local image recognition method based on AR smart glasses.

[0051] Figure 2 It is a block diagram of the specific process steps for generating a moving object marker in a local image recognition method based on an AR smart glasses.

[0052] Figure 3 It is a block diagram of the specific process steps for generating feature information in a local image recognition method based on an AR smart glasses.

[0053] Figure 4 It is a block diagram of the specific process steps for outputting a tracking instruction in a local image recognition method based on an AR smart glasses.

[0054] Figure 5 It is a block diagram of the structure of a local image recognition system based on an AR smart glasses.

[0055] Figure 6 It is a block diagram of the structure of the global construction module in a local image recognition system based on an AR smart glasses.

[0056] Figure 7 It is a block diagram of the structure of the sampling and recognition module in a local image recognition system based on an AR smart glasses. Detailed implementation manners

[0057] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0058] The following describes in detail the specific implementation manners of the present invention with reference to specific embodiments.

[0059] As Figure 1 described, a local image recognition method based on an AR smart glasses provided by an embodiment of the present invention includes the following steps:

[0060] S200, obtaining dynamic image information within the field of view of the device in real time, and performing three-dimensional motion feature analysis on the image information to generate a moving object marker.

[0061] S400, performing dynamic image acquisition on the moving object marker and extracting the feature information of the moving object marker.

[0062] S600, performing cross-comparative analysis on the feature information of the moving object marker and preset object information to generate a tracking instruction.

[0063] S800, sending the tracking instruction to a collaborative device through an intranet and tracking the moving object.

[0064] In an embodiment of the present invention, in step S200, the device here refers to a wearable portable AR device such as an AR glasses. Here, the AR glasses are taken as an example for illustration; the scenario in which this method is used and implemented is at least one AR glasses with the ability to execute this method. Using multiple AR devices simultaneously can obtain better usage effects. When relevant personnel are using it, they wear the AR glasses to observe the surrounding environment, search for the appearance of the target person or object, and perform task tracking. At this time, the worn AR glasses will continuously capture and collect images in the line-of-sight direction. By processing the collected images, moving objects in the images are judged and marked to facilitate the collection and recognition of local images. Then, in step S400, the AR glasses will perform local high-definition image collection on all marked moving objects (i.e., the moving object marks described in the method), and perform feature analysis through the collected images to extract the feature information of the moving object. After obtaining the feature information, in step S600, the feature information is compared and identified with the feature information of the person to be tracked preset by the relevant personnel. If they are the same, it means that the tracked person has been retrieved, and then continuous tracking is performed to prevent the target from being lost; there is a step of sending a tracking instruction to a collaborative device in S800. The collaborative device here refers to the AR devices carried by other relevant personnel in the same duty task. Through data exchange among multiple devices, the device group can cover a larger scanning range, improve the success rate of tracking the person to be tracked, and also reduce the occurrence of the situation of target loss during the tracking process to a certain extent. Even if the target tracking of one device is lost, the relative geographical position of the target can still be obtained through other collaborative devices.

[0065] As another preferred embodiment of the present invention, the dynamic image information is composed of a plurality of consecutive static images collected at a preset time interval, and the preset time interval is used to characterize the movement angular velocity of the device's field of view. The preset time interval is set in an inverse relationship with the movement angular velocity of the device's field of view.

[0066] In the embodiment of the present invention, the dynamic image information in step S200 is explained. Here, the dynamic image refers to a set of static images continuously acquired at a preset time interval, and the preset time interval here is not a constant time interval. A speed sensor is provided in the device, which can sense the rotational angular velocity of the device, that is, the AR glasses, relative to the wearer's neck on the wearer's head. When the wearer's head rotates slowly or is relatively stationary, the environmental change speed in the real - environment picture collected by the AR glasses is relatively slow. At this time, the preset interval time is longer, which can reduce the operating pressure of the device, improve the detection accuracy, and better identify the moving - object marks collected. When the wearer turns the head extremely fast, because the real - environment picture will quickly pass through the visible range of the AR glasses, a shorter preset interval time is required at this time to allow the AR glasses to collect more static images to form dynamic image information and complete the establishment of the three - dimensional space model in the subsequent steps.

[0067] As Figure 2 shown, as another preferred embodiment of the present invention, the dynamic image information further includes distance data and device - position information. The distance data is used to represent the straight - line distance between each image point on the dynamic image data and the device. The step of performing three - dimensional motion - feature analysis on the image information to generate moving - object marks specifically includes:

[0068] S202, record and store the device - position information in real time, establish a three - dimensional space model, and generate a device motion trajectory.

[0069] S203, read the static - image information in the dynamic image information frame by frame, and obtain the distance information and device - position information in the static - image information.

[0070] S204, establish a real - time object model in the three - dimensional space model according to the distance information in the static - image information corresponding to the device - position information, and store it.

[0071] S205, generate an object - model motion trajectory according to the continuously updated real - time object model.

[0072] S206, perform moving - object marking on the object model in the motion state according to the object - model motion trajectory.

[0073] In the embodiment of the present invention, it is a specific expansion and explanation of step S200. In step S200, a three-dimensional space model will be established by itself, and then a three-dimensional space model will be established according to the distance data and position information in the collected dynamic image information in combination with the movement trajectory information of the device itself. Here, the position information represents the angular information relative to the front view direction of the AR glasses, and the distance data refers to the distance from the AR glasses; with the real-time updated dynamic image information, a general description of the surrounding environment will be made in the three-dimensional space model, and then there will be objects relatively stationary in space and objects constantly moving. Then, the step marks the relatively moving objects to facilitate local sampling in subsequent steps for feature extraction, and the stationary objects are used as environmental content without processing.

[0074] As Figure 3 shown, as another preferred embodiment of the present invention, the step of performing dynamic image acquisition on the moving object marker and extracting the feature information of the moving object marker specifically includes:

[0075] S401, obtaining the relative position and distance data of the moving object marker according to the real-time object model.

[0076] S402, performing local image high-definition real-time sampling according to the relative position and distance data of the moving object marker to obtain a local dynamic high-definition image.

[0077] S403, performing feature extraction on the local dynamic high-definition image to generate the feature information of the moving marker body.

[0078] In the embodiment of the present invention, the step of extracting feature information is described in detail. When an object is marked as a moving object, it means that multiple groups of different relative position information and distance information of the object have been collected at this time. Based on these multiple groups of data, its relative position relative to the AR glasses can be determined. At this time, the mechanism with high-definition image acquisition function set on the AR glasses will perform positioning, magnification, and focusing based on this as a reference to obtain a high-definition image of the object, that is, a high-definition image of a local position in the entire view of the AR glasses. Then, by performing feature extraction processing on the image, the feature information of the moving marker body can be generated.

[0079] As Figure 4 shown, as another preferred embodiment of the present invention, the method presets preset object information, and the preset object information is the feature information of the tracking object collected in advance. The step of cross-comparing and analyzing the feature information of the moving object marker with the preset object information to generate a tracking instruction specifically includes:

[0080] S601, establishing a feature comparison model according to the preset object information, and the feature comparison model is used to represent the degree of coincidence with the preset object information.

[0081] S602. Obtain the feature information of the moving object marker, and perform comparison and analysis on the feature information of the moving object marker according to the feature comparison model to generate a comparison result.

[0082] S603. If the comparison result is highly coincident, generate a prompt message and output a tracking instruction as tracking.

[0083] In the embodiment of the present invention, the process of performing comparison and analysis on the feature information is described. This technology is based on the prior art. After the comparative analysis is completed, if it is a tracking target, the wearer is informed by flashing the AR glasses display screen or voice prompt for reminder, and then the object is continuously tracked.

[0084] As another preferred embodiment of the present invention, the step of sending a tracking instruction to the collaborative device through the intranet and tracking the moving object specifically includes:

[0085] When the tracking instruction is tracking, extract the corresponding moving object marker and the tracking instruction.

[0086] Send the moving object marker and the tracking instruction to the collaborative device.

[0087] Perform continuous tracking in the three-dimensional space model according to the corresponding moving marker body information.

[0088] Specifically, the method further includes:

[0089] Receive the moving object marker and the tracking instruction from the collaborative device.

[0090] Retrieve and identify the moving object marker in the three-dimensional space model according to the received moving object marker.

[0091] If the moving object marker in the three-dimensional space model is consistent with the received moving object marker, perform continuous tracking in the three-dimensional space model.

[0092] In the embodiment of the present invention, a simple step description of the collaboration method between multiple devices is provided. That is, when a certain device detects a tracking object, it will make another device also perform synchronous tracking on the object through data exchange, which can greatly improve the tracking efficiency, and the collaboration of multiple devices can increase the field of view coverage and prevent the loss of the target.

[0093] As Figure 5 shown, the present invention also provides a local image recognition system based on an AR smart glasses, which includes:

[0094] S100. A global construction module for real-time obtaining the dynamic image information within the field of view of the device, and performing motion feature analysis on the image information to generate a moving object marker.

[0095] S300, the sampling and recognition module, is used to perform high-definition enlarged sampling on the local image according to the moving object marker, and generate the feature information of the moving object marker.

[0096] S500, the tracking and recognition module, is used to perform cross-comparative analysis on the feature information of the moving object marker and the preset object information, and generate a tracking instruction.

[0097] S700, the collaborative tracking module, is used to send the tracking instruction to the collaborative device through the intranet and track the moving object.

[0098] As another preferred embodiment of the present invention, the global construction module specifically includes:

[0099] S101, the environmental image acquisition unit, is used to acquire the dynamic image information within the field of view of the device in real time.

[0100] S102, the spatial motion model unit, is used to record and store the device position information in real time, establish a three-dimensional space model, and generate the device motion trajectory; read the static image information in the dynamic image information frame by frame, and obtain the distance information and device position information in the static image information; establish a real-time object model in the three-dimensional space model according to the distance information in the static image information corresponding to the device position information and store it; generate the object model motion trajectory according to the continuously updated real-time object model.

[0101] S103, the moving object recognition unit, is used to perform moving object marking on the object model in the moving state according to the object model motion trajectory.

[0102] In the embodiment of the present invention, the environmental image acquisition unit here includes several distance sensors, angle sensors, wide-angle image acquisition lenses with unchangeable focal lengths, and CMOSs, etc.

[0103] As yet another preferred embodiment of the present invention, the sampling and recognition module specifically includes:

[0104] S301, the individual relative positioning module, is used to obtain the relative position and distance data of the moving object marker according to the real-time object model.

[0105] S302, the individual image acquisition unit, is used to perform high-definition real-time sampling of the local image according to the relative position and distance data of the moving object marker, and obtain the local dynamic high-definition image.

[0106] S303, the individual feature acquisition unit, is used to extract the features of the local dynamic high-definition image and generate the feature information of the moving marker body.

[0107] In the embodiment of the present invention, the individual image acquisition unit here includes an image acquisition lens with variable focal length and adjustable focus, and CMOSs, etc.

[0108] It should be understood that although the steps in the flowcharts of the embodiments of the present invention are shown in sequence according to the indication of the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in each embodiment may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same moment, but can be executed at different moments. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or sub-steps or stages of other steps.

[0109] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0110] After considering the specification and the disclosure of the embodiments, those skilled in the art will readily think of other embodiments of the present disclosure. The present application aims to cover any variations, uses, or adaptations of the present disclosure. These variations, uses, or adaptations follow the general principles of the present disclosure and include common general knowledge or conventional technical means in the technical field not disclosed in the present disclosure. The specification and the embodiments are only regarded as exemplary, and the true scope and spirit of the present disclosure are pointed out by the claims.

[0111] It should be understood that the present disclosure is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present disclosure is only limited by the appended claims.

Claims

1. A local image recognition method based on an AR smart glasses, characterized in that, It includes the following steps: Obtain dynamic image information within the field of view of the AR glasses in real time, perform three-dimensional motion feature analysis on the image information, and generate a moving object marker; Perform dynamic image acquisition on the moving object marker and extract the feature information of the moving object marker; Perform cross-comparative analysis on the feature information of the moving object marker and the preset object information to generate a tracking instruction; Send the tracking instruction to the collaborative AR glasses through the intranet and track the moving object; The dynamic image information is composed of several consecutive static images collected at a preset time interval, and the preset time interval is used to represent the motion angular velocity of the field of view of the AR glasses. The preset time interval is set in an inverse relationship with the motion angular velocity of the field of view of the AR glasses; The dynamic image information also includes distance data and position information, and the distance data is used to represent the straight-line distance from each image point on the dynamic image data to the AR glasses; The position information represents the angular information of the actual position of each image point on the dynamic image data relative to the front-facing direction of the AR glasses. The step of performing three-dimensional motion feature analysis on the image information to generate a moving object marker specifically includes: Record and store the position coordinates of the AR glasses in real time, establish a three-dimensional space model, and generate the motion trajectory of the AR glasses; Read the static image information in the dynamic image information frame by frame and obtain the distance data and position information in the static image information; Establish a real-time object model in the three-dimensional space model according to the distance data in the static image information corresponding to the position information and store it; Generate the object model motion trajectory according to the continuously updated real-time object model; Perform a moving object marker on the object model in a moving state according to the object model motion trajectory; The step of performing dynamic image acquisition on the moving object marker and extracting the feature information of the moving object marker specifically includes: Obtain the position information and distance data of the moving object marker according to the real-time object model; Perform local image high-definition real-time sampling according to the position information and distance data of the moving object marker to obtain a local dynamic high-definition image; Extract features from the local dynamic high-definition image to generate the feature information of the moving marker body.

2. The local image recognition method based on an AR smart glasses according to claim 1, wherein The method presets preset object information, and the preset object information is the feature information of the tracking object collected in advance. The step of performing cross-comparative analysis on the feature information of the moving object marker and the preset object information to generate a tracking instruction specifically includes: Establish a feature comparison model according to the preset object information, and the feature comparison model is used to represent the coincidence degree with the preset object information; Obtain the feature information of the moving object marker, and perform comparison analysis on the feature information of the moving object marker according to the feature comparison model to generate a comparison result. If The comparison result is a high degree of coincidence, then generate a prompt message and output the tracking instruction as tracking.

3. The local image recognition method based on an AR smart glasses according to claim 2, wherein The step of sending the tracking instruction to the collaborative AR glasses through the intranet and tracking the moving object specifically includes: When the tracking instruction is tracking, extract the corresponding moving object marker and the tracking instruction; Send the moving object marker and the tracking instruction to the collaborative AR glasses; Perform continuous tracking in the three-dimensional space model according to the corresponding moving marker body information.

4. The local image recognition method based on an AR smart glasses according to claim 1, characterized in that The method further includes: Receiving a moving object marker and tracking instruction from a collaborative AR glasses; Retrieving and identifying the moving object marker in the three-dimensional space model according to the received moving object marker; if The moving object marker in the three-dimensional space model is consistent with the received moving object marker, continuous tracking is performed in the three-dimensional space model.

5. A local image recognition system based on an AR smart glasses, which is used to execute a local image recognition method based on an AR smart glasses as described in claim 1, characterized in that, The local image recognition system specifically includes: A global construction module, configured to obtain dynamic image information within the field of view of the AR glasses in real time, analyze the motion characteristics of the image information, and generate a moving object marker; A sampling recognition module, configured to perform high-definition enlarged sampling on the local image according to the moving object marker to generate feature information of the moving object marker; A tracking recognition module, configured to perform cross-comparison analysis on the feature information of the moving object marker and preset object information to generate a tracking instruction; A collaborative tracking module, configured to send the tracking instruction to the collaborative AR glasses through an intranet and track the moving object.

6. The local image recognition system based on an AR smart glasses according to claim 5, characterized in that, The global construction module specifically includes: An environmental image acquisition unit, configured to obtain dynamic image information within the field of view of the AR glasses in real time; A spatial motion model unit, configured to record and store the position coordinates of the AR glasses in real time, establish a three-dimensional space model, and generate a motion trajectory of the AR glasses; read the static image information in the dynamic image information frame by frame, and obtain the distance data and position information in the static image information; establish a real-time object model in the three-dimensional space model according to the distance data in the static image information corresponding to the position information and store it; generate an object model motion trajectory according to the continuously updated real-time object model; A moving object recognition unit, configured to perform a moving object marker on the object model in a moving state according to the object model motion trajectory.

7. The local image recognition system based on an AR smart glasses according to claim 6, wherein The sampling recognition module specifically includes: An individual relative positioning module, configured to obtain the position information and distance data of the moving object marker according to the real-time object model; An individual image acquisition unit, configured to perform high-definition real-time sampling of the local image according to the position information and distance data of the moving object marker to obtain a local dynamic high-definition image; An individual feature acquisition unit, configured to extract features from the local dynamic high-definition image to generate feature information of the moving marker body.

Images