Target automatic follow-up shooting method and system based on bionic eagle eye

Abstract

The present application relates to a kind of target automatic follow-up shooting method and system based on bionic eagle eye, the method includes deploying several bionic eagle eye cameras in target site, the camera integrates double-axis turntable and wide-angle, long focus, far infrared camera;Hot source label and image label are deployed on the target to be followed up.By far infrared camera tracking hot source label to determine its first pixel coordinates in infrared image, and based on homography matrix, the coordinates are mapped to visible light image, obtain second pixel coordinates.By generating target queue to be followed up and in turn dispatch, drive double-axis turntable rotation makes long focus camera optical axis align target, and then identify image label and obtain identity information.According to the preset anti-repeated shooting logic, system only controls long focus camera to track and shoot when target is not recently shot, and video data is associated with identity information and is stored.The present application realizes the efficient, accurate and automatic follow-up shooting service to moving target in the scene where target feature is not obvious.

Description

Technical Field

[0001] This invention relates to the field of automatic tracking photography technology, and in particular to a method and system for automatic target tracking photography based on a biomimetic eagle eye. Background Technology

[0002] In skiing, both skiers and ski resort operators have a need to record and film the skiing process and exciting moments. With technological advancements, ski resorts have developed various filming systems to meet these needs. However, existing technologies still have significant limitations in practical applications, mainly in the following two aspects:

[0003] First, existing systems are highly dependent on manual labor and have a low degree of automation. Many ski resorts still use the traditional "fixed-point camera + manual supervision" model, or rely on photographers to use handheld equipment for follow-up shooting. This model is not only costly in terms of manpower, but also limited by the photographer's attention and physical strength, making it difficult to cover a large area of ​​ski slopes at the same time, and it cannot guarantee that every tourist can be photographed in an undifferentiated manner, which easily leads to missed shots or untimely follow-up shooting.

[0004] Secondly, existing automatic shooting systems lack sufficient intelligence and effective target identification and tracking capabilities. Some automatic shooting solutions only employ simple long-duration recording methods, which generate massive amounts of invalid video data, resulting in extremely high costs for post-production screening and editing. More advanced solutions attempt to introduce computer vision-based image processing methods, combining camera shooting with facial recognition technology to capture specific objects.

[0005] However, the unique characteristics of skiing scenarios pose challenges to the aforementioned vision-based object detection methods. Skiers typically wear heavy ski suits, and the rental suits provided by ski resorts are often uniform in style and color (such as uniform red or blue), resulting in minimal differences in appearance between individuals and a lack of texture features that algorithms can differentiate. Furthermore, the goggles, helmets, and masks worn by skiers severely obscure facial features, rendering traditional algorithms that rely on facial feature points for recognition virtually ineffective.

[0006] Therefore, how to effectively identify and track moving targets in places such as ski resorts where target features are not obvious, and how to photograph or record them to output high-quality videos or images, is a problem that urgently needs to be solved. Summary of the Invention

[0007] The purpose of this invention is to provide a target automatic tracking and shooting method and system based on bionic eagle eye, so as to solve the problems mentioned in the background art, such as the difficulty in effectively identifying and tracking moving targets for shooting in places such as ski resorts where the target features are not obvious.

[0008] To achieve the above objectives, this invention provides a method for automatic target tracking based on a biomimetic eagle eye, comprising the following steps:

[0009] S1. Deploy several biomimetic eagle-eye cameras at the target site. The biomimetic eagle-eye cameras include a dual-axis turntable and a wide-angle camera, a far-infrared camera and a telephoto camera mounted on the dual-axis turntable.

[0010] S2. Deploy heat source tags and image tags on all targets to be tracked;

[0011] S3. The infrared image and visible light image of the target site are acquired by the far-infrared camera and the wide-angle camera respectively, the heat source tag deployed on the target to be tracked is identified, and the first pixel coordinate of the heat source tag in the infrared image is determined.

[0012] S4. Based on the homography matrix between the far-infrared image and the visible light image obtained through pre-calibration, the first pixel coordinates are mapped to the visible light image to obtain the second pixel coordinates of the heat source tag in the visible light image;

[0013] S5. Select and number the second pixel coordinates of multiple targets to be followed in sequence to generate a target queue;

[0014] S6. Generate a control command based on the second pixel coordinates corresponding to the current number in the target queue to drive the dual-axis turntable to rotate so as to align the optical axis center of the telephoto camera with the physical orientation corresponding to the second pixel coordinates.

[0015] S7. Analyze the image located in the center area of ​​the telephoto camera's field of view, identify the image tag, and obtain the target's identity information; if no valid image tag is identified, return to execute S6 and generate a control command based on the second pixel coordinates corresponding to the next number in the target queue to be followed; otherwise, proceed to S8.

[0016] S8. Based on the preset anti-repeated shooting logic, determine whether the target identity information has been recorded as already shot within a preset time period; if yes, return to execute S6 and generate a control command based on the second pixel coordinates corresponding to the next number in the target queue; if no, control the telephoto camera to track and shoot the target, and associate and store the generated tracking video data with the target identity information.

[0017] Optionally, the heat source tag is a tag whose surface can emit heat.

[0018] Optionally, the image label may employ at least one of pattern-based or color-based distinguishing identifiers.

[0019] Optionally, the heat source label and the image label can be designed separately or as a single unit.

[0020] Optionally, the step of associating and storing the generated follow-up video data with the target's identity information specifically includes:

[0021] The tracking video data is classified according to the image tags, and the classified tracking video data is transmitted to the background storage system. The background storage system receives all the tracking video data from the bionic eagle eye camera.

[0022] The background storage system is used to set tag IDs for all image tags, and all follow-up video data is categorized and stored under the corresponding tag IDs.

[0023] Optionally, the preset time period is a cooling-off time window from the moment the last shooting was completed, and the anti-repeated shooting logic is used to avoid repeatedly shooting the same target within the cooling-off time window.

[0024] Based on the same inventive concept, the present invention also provides a system applied to the above-described biomimetic eagle eye-based automatic target tracking method, comprising:

[0025] The shooting system includes several bionic eagle-eye cameras;

[0026] Tracking tags include heat source tags and image tags deployed on the target to be tracked;

[0027] An image processing system is installed on each of the bionic eagle-eye cameras. The image processing system is equipped with an algorithm for detecting heat source tags in far-infrared images, an image processing algorithm for identification and localization, an alignment algorithm for calculating the alignment between the far-infrared image and the visible light image, an algorithm for driving and controlling the dual-axis turntable, and an algorithm for identifying and detecting image tags in the visible light image.

[0028] The background storage system is connected in communication with the image processing system and is used to associate and store the generated follow-up video data with the target's identity information.

[0029] Optionally, the dual-axis turntable includes a pitch axis turntable and a yaw axis turntable, the pitch axis turntable is mounted on the yaw axis turntable, and a wide-angle camera, a far-infrared camera and a telephoto camera are mounted on the pitch axis turntable;

[0030] The yaw axis turntable is used to realize the rotation of the pitch axis turntable and the wide-angle camera, far-infrared camera and telephoto camera mounted on it in the horizontal plane.

[0031] The pitch axis turntable is used to rotate the wide-angle camera, far-infrared camera, and telephoto camera in a vertical plane.

[0032] Optionally, the bionic eagle-eye camera includes a motion controller and a motor controlled by the motion controller. The motor is connected to the dual-axis turntable and is used to drive the dual-axis turntable to rotate.

[0033] The target automatic tracking and shooting method and system based on bionic eagle eye provided by this invention has at least one of the following beneficial effects:

[0034] 1) This invention utilizes a dual verification mechanism of heat source tags and image tags, employing a far-infrared camera to penetrate the visual blind spots caused by uniform ski suits and facial occlusion, quickly locking onto and distinguishing targets that need to be photographed from those that do not, thus solving the failure problem of traditional face recognition-based solutions in scenarios with indistinct features; at the same time, it uses a telephoto camera to accurately identify image tags, realizing the automatic binding of personnel codes and identity information, providing a precise data foundation for the subsequent classification, storage, retrieval, and management of video data.

[0035] 2) Compared with the expensive tag solutions that rely on GPS or radio transceivers, the thermal source tags and image tags used in this invention have significant advantages such as simple manufacturing, low cost, and no need for power supply. In addition, the thermal source tags have extremely high thermal radiation contrast in low temperature environments, which not only greatly reduces hardware deployment and maintenance costs, but also avoids the impact of electromagnetic interference on positioning accuracy.

[0036] 3) A hardware architecture combining wide-angle camera sensing, far-infrared camera precise positioning, and telephoto camera close-up shooting, along with a pre-calibrated homography matrix, enables precise mapping of heterogeneous coordinates. A dual-axis turntable (pitch and yaw axes) then rapidly adjusts the optical axis orientation based on the mapped coordinates. This deeply integrated hardware and software solution ensures at the hardware level that the system can quickly and accurately track high-speed moving targets, replacing the inefficient traditional manual or semi-automatic shooting modes and achieving unattended, fully automatic tracking. Attached Figure Description

[0037] Those skilled in the art will understand that the accompanying drawings are provided to better understand the invention and do not constitute any limitation on the scope of the invention. Wherein:

[0038] Figure 1 A flowchart of a target automatic tracking and shooting method based on a biomimetic eagle eye provided in an embodiment of the present invention;

[0039] Figure 2 This is a schematic diagram of the structure of a biomimetic eagle-eye camera provided in an embodiment of the present invention;

[0040] Figure 3 This is a schematic diagram of the deployment of a shooting system provided in an embodiment of the present invention at a ski resort.

[0041] The attached figures are labeled as follows:

[0042] 1-Wide-angle camera; 2-Far-infrared camera; 3-Telephoto camera; 4-Pitch axis turntable; 5-Yaw axis turntable; 6-Camera mounting interface. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0044] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0045] In the description of this invention, it should be understood that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product is in use, or the orientation or positional relationship commonly understood by those skilled in the art. They are only used to facilitate the description of this invention and to simplify the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0046] Furthermore, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such an article or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the article or apparatus that includes said element. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0047] On the one hand, such as Figure 1 and Figure 2As shown, this embodiment of the invention provides a method for automatic target tracking based on a biomimetic eagle eye, including the following steps:

[0048] S1. Deploy several biomimetic eagle-eye cameras at the target site. The biomimetic eagle-eye cameras include a dual-axis turntable and a wide-angle camera, a far-infrared camera and a telephoto camera mounted on the dual-axis turntable.

[0049] S2. Deploy heat source tags and image tags on all targets to be tracked;

[0050] S3. The infrared image and visible light image of the target site are acquired by the far-infrared camera and the wide-angle camera respectively, the heat source tag deployed on the target to be tracked is identified, and the first pixel coordinate of the heat source tag in the infrared image is determined.

[0051] S4. Based on the homography matrix between the far-infrared image and the visible light image obtained through pre-calibration, the first pixel coordinates are mapped to the visible light image to obtain the second pixel coordinates of the heat source tag in the visible light image;

[0052] S5. Select and number the second pixel coordinates of multiple targets to be followed in sequence to generate a target queue;

[0053] S6. Generate a control command based on the second pixel coordinates corresponding to the current number in the target queue to drive the dual-axis turntable to rotate so as to align the optical axis center of the telephoto camera with the physical orientation corresponding to the second pixel coordinates.

[0054] S7. Analyze the image located in the center area of ​​the telephoto camera's field of view, identify the image tag, and obtain the target's identity information; if no valid image tag is identified, return to execute S6 and generate a control command based on the second pixel coordinates corresponding to the next number in the target queue to be followed; otherwise, proceed to S8.

[0055] S8. Based on the preset anti-repeated shooting logic, determine whether the target identity information has been recorded as already shot within a preset time period; if yes, return to execute S6 and generate a control command based on the second pixel coordinates corresponding to the next number in the target queue; if no, control the telephoto camera to track and shoot the target, and associate and store the generated tracking video data with the target identity information.

[0056] On the other hand, such as Figure 2 As shown, this embodiment of the invention provides a system for the above-mentioned biomimetic eagle eye-based automatic target tracking method, comprising:

[0057] The shooting system includes several bionic eagle-eye cameras;

[0058] Tracking tags include heat source tags and image tags deployed on the target to be tracked;

[0059] An image processing system, installed on each bionic eagle-eye camera, is equipped with algorithms for detecting heat source tags in far-infrared images, image processing algorithms for identification and localization, alignment algorithms for calculating the alignment between far-infrared and visible light images, algorithms for driving and controlling the dual-axis turntable, and algorithms for identifying and detecting image tags in visible light images.

[0060] The back-end storage system communicates with the image processing system and is used to associate and store the generated follow-up video data with the target's identity information.

[0061] This configuration, utilizing a dual verification mechanism of thermal source tags and image tags, enables the far-infrared camera 2 to penetrate visual blind spots caused by uniform ski clothing and facial occlusion in low-temperature environments such as ski resorts. This accurately locks onto and distinguishes targets requiring filming, overcoming the failure issues of traditional vision-based or facial recognition solutions in scenarios with indistinct features. Simultaneously, by using a pre-calibrated homography matrix to losslessly map infrared image coordinates to visible light image coordinates, coupled with the rapid response of a dual-axis turntable, a closed-loop execution process from "target detection" to "long-range locking" is achieved, ensuring both speed and accuracy in tracking. Furthermore, the system achieves automated polling and scheduling of multiple targets through dynamic generation of a target queue and anti-duplication shooting logic, avoiding resource waste. The target identity information-based association storage mechanism allows the generated video data to be automatically bound to personnel information, significantly reducing the cost of manual screening and editing in post-production. Moreover, by employing low-cost, easily deployable thermal source tags, it replaces expensive and easily interfered-with radio positioning solutions such as GPS, achieving a low-cost, high-efficiency, fully automated tracking service.

[0062] In this embodiment, several bionic eagle-eye cameras are installed at appropriate locations around the target site, such as a ski resort or ice rink, forming a shooting system. These bionic eagle-eye cameras do not operate independently but constitute a distributed shooting network. The number and position of the cameras can be flexibly adjusted according to the size of the site and specific needs, ensuring that all activity areas within the site are within the field of view of at least one camera. Figure 2As shown, the bionic eagle-eye camera includes a dual-axis turntable (capable of at least two-dimensional rotation) and a wide-angle camera 1, a far-infrared camera 2, and a telephoto camera 3 mounted on the turntable. The dual-axis turntable includes a pitch-axis turntable 4 and a yaw-axis turntable 5. The pitch-axis turntable 4 is mounted on the yaw-axis turntable 5, and the wide-angle camera 1, far-infrared camera 2, and telephoto camera 3 are mounted on the pitch-axis turntable 4. It also integrates other functional modules such as power, control, and transmission. Furthermore, the pitch-axis turntable 4 is designed with a camera mounting interface 6, which can be used to detachably mount other high-definition shooting equipment to adapt to different shooting needs.

[0063] The yaw axis turntable 5 is used to realize the rotation of the pitch axis turntable 4 and the wide-angle camera 1, far-infrared camera 2 and telephoto camera 3 mounted on it on the horizontal plane.

[0064] The pitch axis turntable 4 is used to realize the rotation of the wide-angle camera 1, the far-infrared camera 2 and the telephoto camera 3 on the vertical plane.

[0065] This structural design enables the yaw axis turntable 5 to rotate 360 ​​degrees on the horizontal plane, thereby covering a wide azimuth angle; the pitch axis turntable 4 is responsible for controlling the pitch movement of the camera group on the vertical plane to capture targets at different heights.

[0066] To achieve precise movement of the turntable, the bionic eagle-eye camera incorporates a motion controller and a motor controlled by that controller. The motor is connected to the dual-axis turntable via a mechanical transmission mechanism (such as gears or belts). Based on received control commands, the motion controller drives the motor to output torque, thereby rotating the dual-axis turntable. The transmission of control signals and image information can utilize wired networks, Wi-Fi, or 4G / 5G, among other methods, to adapt to different field environments.

[0067] In addition, the tilt axis turntable 4 is designed with a universal camera mounting interface 6. This interface supports standard threaded or snap-fit ​​connections for the detachable mounting of additional high-definition shooting equipment. This means that in addition to the standard wide-angle, telephoto, and infrared cameras, users can add other professional video equipment according to their actual needs (such as the need for higher resolution or special perspective shooting), enhancing the system's expandability and compatibility.

[0068] A tracking tag needs to be deployed on the target to be filmed (such as a skier). The tracking tag consists of two parts: a heat source tag and an image tag. The heat source tag is key to the rapid target detection in complex backgrounds. It is a tag whose surface can emit heat, such as a patch containing a chemically heated material or a small electric heating element. In low-temperature environments such as ski resorts, the heat radiation emitted by this tag contrasts sharply with the low temperature of the surrounding snow and ice, making it extremely easy for the far-infrared camera 2 to identify it without being affected by snow reflections or the heavy clothing worn by the skier.

[0069] Image tags are used to provide specific identification information. They take at least one form of identification, such as pattern identification or color identification. For example, they can be stickers printed with special geometric patterns, QR codes, or barcodes, or they can be identification objects with bright color contrast. Heat source tags and image tags can be designed separately and applied to different parts of the visitor's body; or they can be designed as an integrated unit, combining both on the same physical carrier for easy wearing and management.

[0070] Each biomimetic eagle-eye camera is equipped with an image processing system containing relevant processing algorithms, including but not limited to algorithms for detecting heat source tags in far-infrared images, image processing algorithms for identification and localization, algorithms for calculating the alignment between far-infrared and visible light images, algorithms for driving and controlling the dual-axis turntable, and algorithms for identifying and detecting image tags in visible light images. It should be noted that the core innovation of this invention lies in the collaborative working method of the aforementioned hardware architecture, rather than in the specific underlying algorithms themselves. Those skilled in the art will understand that the following algorithms are existing technologies in the fields of machine vision and automatic control, which can be directly invoked according to the actual hardware configuration or implemented through conventional programming, without requiring any creative effort.

[0071] The system based on the bionic eagle eye target automatic tracking method also includes a background storage system that communicates with the image processing system. It can automatically sort, classify and summarize data from multiple bionic eagle eye cameras. In the background storage system, each tag ID can correspond to the tracking video data captured by different eagle eye devices.

[0072] This biomimetic eagle-eye-based automatic target tracking system is suitable for various sports and venues, such as indoor and outdoor skiing, skating, swimming, mountain cable cars, skydiving, and paragliding, where it is suitable for tracking and filming large groups of people participating in sports and recreational activities. This embodiment takes a ski resort as an example, and the specific process of its automatic target tracking method is as follows:

[0073] First, execute S1, deploying several bionic eagle-eye cameras at the target site, such as... Figure 3As shown, A, B, C, and D represent multiple bionic eagle-eye cameras deployed at optimal shooting locations near the ski resort. Each bionic eagle-eye camera simultaneously acquires image data of the target site via a far-infrared camera 2 and a wide-angle camera 1. The far-infrared camera 2 captures infrared images and utilizes its thermal imaging characteristics to detect all heat source tags in the image through image processing techniques. The system determines the first pixel coordinates of each heat source tag in the infrared image. Simultaneously, the wide-angle camera 1 acquires visible light images for subsequent color and texture analysis.

[0074] Then, step S2 is executed to deploy heat source tags and image tags on all targets to be filmed. In this embodiment, heat source tags and image tags are deployed on targets to be filmed (such as paying tourists, athletes training, etc.) before skiing activities begin.

[0075] Next, S3 is executed, which uses far-infrared camera 2 and wide-angle camera 1 to acquire infrared and visible light images of the target site, respectively, to identify the heat source tag deployed on the target to be tracked, and to determine the first pixel coordinates of the heat source tag in the infrared image.

[0076] During the skiing activity, each of the system's bionic eagle-eye cameras starts working. Taking bionic eagle-eye camera B as an example, the far-infrared camera 2 mounted on camera B, together with its own far-infrared image heat source tag detection algorithm, detects all heat source tags in the field in the far-infrared image, and uses the image processing algorithm for recognition and positioning to determine the first pixel coordinates of each heat source tag in the far-infrared image.

[0077] Next, S4 is executed. Based on the homography matrix between the far-infrared image and the visible light image obtained by pre-calibration, the first pixel coordinates are mapped to the visible light image to obtain the second pixel coordinates of the heat source tag in the visible light image.

[0078] Because the optical centers of the far-infrared camera 2 and the wide-angle camera 1 do not coincide, and perspective distortion may exist, the coordinates in the infrared image cannot be directly used to control the visible light camera. Therefore, the system utilizes a pre-calibrated homography matrix. This matrix describes the projection transformation relationship between the infrared image plane and the visible light image plane. By substituting the first pixel coordinates into this matrix for calculation, the system can obtain the second pixel coordinates of the heat source tag in the visible light image. This process achieves spatial alignment of data from different sensor sources.

[0079] Then execute S5 to select and number the second pixel coordinates of multiple targets to be followed, generating a target queue.

[0080] In this embodiment, the system summarizes the second pixel coordinates of multiple heat source tags detected within the field of view, and selects and numbers them sequentially according to certain rules (such as from left to right, or from near to far). Figure 3 The system uses ①, ②, and ③ to generate an ordered queue of targets to be tracked. This queue management mechanism ensures that the system can provide fair and orderly services to each tagged target.

[0081] Next, S6 is executed, generating a control command based on the second pixel coordinates corresponding to the current number (e.g., number ①) in the target queue, driving the dual-axis turntable to rotate so that the optical axis center of the telephoto camera 3 is aligned with the physical orientation corresponding to the second pixel coordinates.

[0082] In this embodiment, the system generates control commands based on the second pixel coordinates corresponding to the current number in the target queue. Upon receiving the command, the motion controller drives the motor to rotate, causing the dual-axis turntable to rotate. The goal of this rotation is to precisely align the optical axis center of the telephoto camera 3 with the physical location corresponding to the second pixel coordinates; that is, to control the dual-axis turntable of the bionic eagle-eye camera to rotate until the second pixel coordinates are located at the center of the field of view of the telephoto camera 3's image. Because the telephoto camera 3 has a narrow field of view, this precise alignment is crucial to ensuring that the target appears in the center of the frame.

[0083] Next, step S7 is executed to analyze the image located in the center of the field of view of the telephoto camera 3, identify image tags, and obtain target identity information. If no valid image tag is identified, the process returns to step S6 to generate a control command based on the second pixel coordinates corresponding to the next number in the target queue; otherwise, it proceeds to step S8.

[0084] In this embodiment, after the telephoto camera 3 is pointed at the target, the system analyzes the image of the central area of ​​the telephoto camera 3's field of view. Using image recognition algorithms (such as QR code decoding or pattern matching algorithms), the system attempts to identify image tags to obtain target identity information (such as tourist ID).

[0085] Finally, S8 is executed, and the target identity information is determined according to the preset anti-repeated shooting logic to determine whether it has been recorded as shot within the preset time period. If yes, it returns to S6 and generates a control command according to the second pixel coordinates corresponding to the next number in the target queue. If no, it controls the telephoto camera 3 to track and shoot the target and associates the generated tracking video data with the target identity information for storage.

[0086] In this embodiment, after obtaining the target's identity information, the system makes a judgment based on a preset anti-duplicate shooting logic. The preset time period is set as a cooldown time window (e.g., 5 minutes) from the time the last shooting was completed. The system checks whether the target's identity information has been recorded as having been shot within this time window.

[0087] If the system determines that the target has already been photographed within the cooling time window, or if the system fails to identify valid image tag information in the center of the field of view (which may be a false detection or a non-service object), the system will abandon the current target, return to S6, and aim the telephoto camera at the second pixel coordinate corresponding to number ②.

[0088] If the system determines that the target has not been photographed before, it will control the telephoto camera 3 to track and photograph the target. During the shooting process, the system can adjust the attitude of the dual-axis turntable in real time according to the movement trajectory of the heat source tag, keeping the target always in the center of the telephoto camera 3's field of view. After tracking for a period of time, the system records the captured state of the target and restarts polling the next target in the queue.

[0089] During filming, the generated follow-up video data is linked and stored in real time with the identified target's identity information. Specifically, the image processing system categorizes the follow-up video data based on image tags and transmits the categorized video data to the backend storage system. The backend storage system receives all video data uploaded by the bionic eagle-eye cameras and assigns a unique tag ID to each image tag. Subsequently, the system stores all follow-up video data under the corresponding tag ID. For example, all video clips of a tourist with tag ID "12345" taken from different camera positions and at different times will be automatically categorized into the "12345" folder. This automated data management method greatly simplifies the subsequent video editing, retrieval, and distribution processes.

[0090] In summary, this invention provides a method and system for automatic target tracking and shooting based on a bionic eagle eye. By controlling the coordination between the three modules of the bionic eagle eye camera—infrared, wide-angle, and telephoto—and the motion control of the dual-axis turntable, it achieves automatic tracking and shooting of moving targets in locations such as ski resorts. Compared to traditional manual or semi-automatic shooting methods, this invention frees up manpower while enabling fully automatic tracking and shooting of the subject. Furthermore, through heat sources and image tags, it can quickly identify personnel codes and match them with personnel information, facilitating the classification and storage of captured images or videos, as well as subsequent retrieval, storage, and management of photos or videos.

[0091] The above description is merely a description of preferred embodiments of the present invention and is not intended to limit the scope of the invention in any way. Any changes or modifications made by those skilled in the art based on the above disclosure are within the protection scope of the present invention. Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the present invention and its equivalents, the present invention also intends to include these modifications and variations.

Claims

1. A method for automatic target tracking based on biomimetic eagle eye, characterized in that, Includes the following steps: S1. Deploy several biomimetic eagle-eye cameras at the target site. The biomimetic eagle-eye cameras include a dual-axis turntable and a wide-angle camera, a far-infrared camera and a telephoto camera mounted on the dual-axis turntable. S2. Deploy heat source tags and image tags on all targets to be tracked; S3. The infrared image and visible light image of the target site are acquired by the far-infrared camera and the wide-angle camera respectively, the heat source tag deployed on the target to be tracked is identified, and the first pixel coordinate of the heat source tag in the infrared image is determined. S4. Based on the homography matrix between the far-infrared image and the visible light image obtained through pre-calibration, the first pixel coordinates are mapped to the visible light image to obtain the second pixel coordinates of the heat source tag in the visible light image; S5. Select and number the second pixel coordinates of multiple targets to be followed in sequence to generate a target queue; S6. Generate a control command based on the second pixel coordinates corresponding to the current number in the target queue to drive the dual-axis turntable to rotate so as to align the optical axis center of the telephoto camera with the physical orientation corresponding to the second pixel coordinates. S7. Analyze the image located in the center area of ​​the telephoto camera's field of view, identify the image tag, and obtain the target's identity information; If no valid image label is identified, return to execute S6 and generate a control command based on the second pixel coordinates corresponding to the next number in the target queue to be followed; Conversely, proceed to S8; S8. Based on the preset anti-repeated shooting logic, determine whether the target identity information has been recorded as a shot within a preset time period; If yes, then return to execute S6 and generate a control command based on the second pixel coordinates corresponding to the next number in the target queue; if no, then control the telephoto camera to track and shoot the target, and associate and store the generated tracking video data with the target identity information.

2. The automatic target tracking method based on bionic eagle eye according to claim 1, characterized in that, The heat source tag is a tag whose surface can emit heat.

3. The automatic target tracking method based on bionic eagle eye according to claim 1, characterized in that, The image label uses at least one of pattern-based or color-based distinguishing identifiers.

4. The automatic target tracking method based on bionic eagle eye according to claim 1, characterized in that, The heat source label and image label can be designed separately or as a single unit.

5. The automatic target tracking method based on bionic eagle eye according to claim 1, characterized in that, The step of associating and storing the generated follow-up video data with the target's identity information specifically includes: The tracking video data is classified according to the image tags, and the classified tracking video data is transmitted to the background storage system. The background storage system receives all the tracking video data from the bionic eagle eye camera. The background storage system is used to set tag IDs for all image tags, and all follow-up video data is categorized and stored under the corresponding tag IDs.

6. The automatic target tracking method based on bionic eagle eye according to claim 1, characterized in that, The preset time period is a cooling-off time window from the moment the last shot was completed. The anti-repeated shooting logic is used to avoid repeatedly shooting the same target within the cooling-off time window.

7. A system for automatically tracking and filming targets based on a biomimetic eagle eye according to any one of claims 1-6, characterized in that, include: The shooting system includes several bionic eagle-eye cameras; Tracking tags include heat source tags and image tags deployed on the target to be tracked; An image processing system is installed on each of the bionic eagle-eye cameras. The image processing system is equipped with an algorithm for detecting heat source tags in far-infrared images, an image processing algorithm for identification and localization, an alignment algorithm for calculating the alignment between the far-infrared image and the visible light image, an algorithm for driving and controlling the dual-axis turntable, and an algorithm for identifying and detecting image tags in the visible light image. The background storage system is connected in communication with the image processing system and is used to associate and store the generated follow-up video data with the target's identity information.

8. The target automatic tracking system based on bionic eagle eye according to claim 7, characterized in that, The dual-axis turntable includes a pitch axis turntable and a yaw axis turntable. The pitch axis turntable is mounted on the yaw axis turntable, and a wide-angle camera, a far-infrared camera, and a telephoto camera are mounted on the pitch axis turntable. The yaw axis turntable is used to realize the rotation of the pitch axis turntable and the wide-angle camera, far-infrared camera and telephoto camera mounted on it in the horizontal plane. The pitch axis turntable is used to rotate the wide-angle camera, far-infrared camera, and telephoto camera in a vertical plane.

9. The automatic target tracking system based on bionic eagle eye according to claim 7, characterized in that, The bionic eagle-eye camera is equipped with a motion controller and a motor controlled by the motion controller. The motor is connected to the dual-axis turntable and is used to drive the dual-axis turntable to rotate.

10. The target automatic tracking system based on bionic eagle eye according to claim 7, characterized in that, The pitch axis turntable is equipped with a camera mounting interface for the detachable mounting of additional high-definition shooting equipment.

Images