Target tracking method and system, electronic equipment and computer readable storage medium

Abstract

The invention relates to the technical field of servers, in particular to a target tracking method and system, electronic equipment and a computer readable storage medium, and the method comprises the steps: starting and controlling a panoramic lens to move left and right, carrying out the target detection of a sports scene, and achieving the global coverage monitoring of the sports scene; identifying a tracking target in the sports scene, and synchronously acquiring position information of the tracking target in the view field range of the panoramic lens to realize accurate positioning of a key tracking target; based on the position information, the close-up lens is controlled to rotate to the position where the tracking target is located, the close-up lens is started to track the tracking target in real time, high-definition and real-time capture of the motion state and motion details of the tracking target is achieved, and then the stability of target tracking in the sports scene is improved.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to a target tracking method, system, electronic device, and computer-readable storage medium. Background Technology

[0002] With the widespread application of computer vision and deep learning technologies in the sports field, PTZ cameras with automatic tracking capabilities have become core equipment in AI sports live streaming, training analysis, and event review scenarios. They need to continuously track targets in large-field events such as football to ensure that key footage of the event is not lost.

[0003] In current processing methods, while multi-view stitching panoramic camera solutions can cover a wide field of view, they require electronic zoom keying to output the image. To ensure clarity, the resolution needs to be increased to 12k or higher, resulting in a surge in hardware and computing costs and insufficient economic efficiency. Although monocular PTZ cameras can adjust the monitoring range, their field of view coverage is relatively narrow at the same time. In events such as football, once the tracked target is lost, it is extremely difficult to relocate the target, making continuous and stable tracking impossible.

[0004] Among the above methods, the core problem of monocular and fixed binocular gimbal solutions lies in insufficient tracking stability, which makes it impossible to continuously track sports targets. This problem directly restricts the reliable application of AI sports tracking technology in real-world scenarios, and the technical problem of insufficient target tracking stability in sports scenarios urgently needs to be solved.

[0005] The above content is only used to help understand the technical solution of this application and does not represent an admission that the above content is prior art. Summary of the Invention

[0006] The main objective of this application is to provide a target tracking method, system, electronic device, and computer-readable storage medium, which aims to improve the stability of target tracking in sports scenarios.

[0007] To achieve the above objectives, this application provides a target tracking method applied to a target tracking system, the system including a binocular pan-tilt camera, the binocular pan-tilt camera including a panoramic lens and a close-up lens, the target tracking method including: Start and control the panoramic camera to move left and right to perform target detection in the sports scene; Identify the tracking target in the sports scene and simultaneously acquire the position information of the tracking target within the field of view of the panoramic lens; Based on the location information, the camera lens is controlled to rotate to the location of the target being tracked, and the camera lens is activated to track the target in real time.

[0008] Furthermore, to achieve the above objectives, this application also provides a target tracking system, which includes a binocular pan-tilt camera comprising a panoramic lens and a close-up lens. The target tracking system includes: The target detection module is used to start and control the left and right movement of the panoramic camera to detect targets in the sports scene; The information acquisition module is used to identify the tracking target in the sports scene and simultaneously acquire the position information of the tracking target within the field of view of the panoramic lens; The lens control module is used to control the close-up lens to rotate to the location of the target based on the position information, and to start the close-up lens to track the target in real time.

[0009] Each functional module of the target tracking system of this application implements the steps of the target tracking method of this application as described above during runtime.

[0010] In addition, to achieve the above objectives, this application also provides an electronic device, which includes a binocular pan-tilt camera, a memory, a processor, and a target tracking program stored in the memory and executable on the processor. When the target tracking program is executed by the processor, it implements the steps of the target tracking method described above.

[0011] In addition, to achieve the above objectives, this application also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a target tracking program, and the target tracking program implements the steps of the target tracking method described above when executed by a processor.

[0012] This application provides a target tracking method. By activating and controlling the left-right movement of a panoramic lens, target detection is performed on a sports scene, achieving full-area coverage monitoring of the sports scene. This ensures that all potential tracking targets within the scene are initially captured, avoiding target omissions due to the limitations of a single lens's field of view, and providing a comprehensive scene foundation for subsequent accurate identification and tracking. The method identifies tracking targets in the sports scene and simultaneously acquires the position information of the tracking targets within the field of view of the panoramic lens, achieving precise positioning of key tracking targets and clarifying their directional coordinates within the scene. This provides accurate positional basis for the directional rotation of subsequent close-up shots, solving the problem of ambiguous positioning of key targets in multi-target scenes. Based on the position information, the close-up shot is controlled to rotate to the location of the tracking target, and real-time tracking of the target is initiated. This achieves high-definition, real-time capture of the target's motion state and action details, balancing tracking continuity with image detail presentation. This meets the dual requirements of target tracking accuracy and visual effects in scenarios such as live sports broadcasts and training data analysis, thereby improving the stability of target tracking in sports scenes. Attached Figure Description

[0013] Figure 1 This is a flowchart illustrating the first embodiment of the target tracking method of this application; Figure 2 This is a schematic diagram of target tracking in a live football match scenario covered by this application; Figure 3 This is a flowchart of the target tracking method involved in this application; Figure 4 This is a schematic diagram of the target tracking system involved in the embodiments of this application; Figure 5 This is a schematic diagram of the structure of the electronic device involved in the embodiments of this application.

[0014] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0015] This application provides a target tracking method, referring to... Figure 1 As shown, Figure 1 This is a flowchart illustrating the first embodiment of the target tracking method of this application.

[0016] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0017] In scenarios such as live sports broadcasts and training data analysis, accurate tracking of moving targets (such as players and event balls) is a core requirement for improving content quality and analysis efficiency. The continuity of target tracking and the ability to capture details directly affect the live broadcast viewing experience and the effectiveness of training data.

[0018] Current methods for target tracking in sports scenarios mostly involve single-lens independent tracking or multi-lens manual switching. These methods have significant drawbacks: single-lens tracking struggles to balance wide-area coverage with detail capture; if a wide-angle lens is used, target details become blurred, while if a telephoto lens is used, the field of view becomes narrow, making it easy to lose fast-moving targets; multi-lens manual switching relies on manual operation, resulting in high switching latency, inability to follow the target's movement trajectory in real time, frequent interruptions in the tracking process, and insufficient accuracy.

[0019] The underlying reason is that existing methods lack a coordinated design for "full coverage" and "focus on details," making it difficult to simultaneously meet the dual requirements of not losing targets and maintaining clear details in dynamic sports scenarios. This problem is particularly prominent in high-speed competitive sports events, which has hindered the development of intelligent tracking in sports scenarios.

[0020] This application provides a target tracking method, system, electronic device, and computer-readable storage medium that can effectively solve the problems of insufficient tracking continuity and accuracy, and provide stable and accurate target tracking support for sports scenarios.

[0021] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, mobile phone, database system, etc., or a device capable of performing the above functions, such as an electronic device. The following description uses an electronic device as an example to illustrate this embodiment and the subsequent embodiments.

[0022] The target tracking method of this application includes the following implementation steps S10 to S30.

[0023] Step S10: Start and control the panoramic camera to move left and right to perform target detection in the sports scene; In this embodiment, as an example, the application scenario can be a live broadcast of a large-scale sports event or a training data analysis scenario for professional athletes.

[0024] Optionally, left and right movement refers to controlling the horizontal displacement of the panoramic lens, including but not limited to reciprocating or directional movement. The movement range adapts to the horizontal width of the sports scene, and the movement speed can be pre-configured (in this embodiment, the value range includes but is not limited to 5° / s-15° / s). Activating and controlling the left and right movement of the panoramic lens means simultaneously triggering the horizontal movement control logic of the lens while activating the panoramic lens image acquisition function. This allows the lens to achieve wider coverage detection of the sports scene through horizontal movement during image acquisition. The two functions are executed collaboratively to ensure that detection and movement are synchronized. Target tracking system: a system used for identifying, locating, and continuously tracking targets in a specific scene; a binocular pan-tilt camera integrates two... Lenses (panoramic and close-up lenses) are devices that allow the pan-tilt unit to rotate and adjust the shooting angle. A panoramic lens refers to a lens mounted on a binocular pan-tilt camera that has a wide preset field of view (in this embodiment, it can be set to, but is not limited to, 120°-180°) and can capture images of a large-scale scene. A close-up lens refers to a lens mounted on a binocular pan-tilt camera that has high-magnification zoom capability and is used for detailed shooting and real-time tracking of specific targets. Sports scenes include sports events (such as football matches and basketball matches), professional training, and other activity scenes where fast-moving targets (such as players and match balls) exist. Target detection refers to the technical process of capturing scene images through lenses and identifying preset tracking objects (such as players and match balls) in the images.

[0025] Optionally, this embodiment is applied to live broadcasts of large-scale football matches. Currently, in live sports broadcasts, target detection mostly uses single-lens shooting. If a regular wide-angle lens is used, although it can cover part of the field, the field of view is limited, making it difficult to completely capture targets in the goal areas at both ends; if a telephoto lens is used, although the clarity is high, the field of view is narrow, making it easy to miss fast-moving players; in this embodiment, such as Figure 2 As shown, Figure 2 This is a schematic diagram illustrating target tracking in a live football match scene covered in this application. The panoramic lens of the binocular PTZ camera is activated. Simultaneously, while activating the lens's image acquisition function, a control signal is sent to the PTZ drive module to control the panoramic lens to move horizontally left and right. The panoramic lens has a preset field of view of 150°, adapting to the width of a standard football field. The movement range is set from the lens angle corresponding to the left sideline to the lens angle corresponding to the right sideline (total movement angle is 120°), and the movement speed is set to 10° / s. The lens pauses for 0.5 seconds every 10° movement to acquire images of the current area, ensuring image clarity. During the left-right movement, the panoramic lens continuously acquires image data of the covered area. Through segmented acquisition and stitching, it covers the entire grass area, player activity area, and key locations such as the sidelines and goals, completely capturing all potential tracking targets (players, football) within the scene, thus completing target detection in the sports scene.

[0026] Optionally, this embodiment is applied to a basketball player training data analysis scenario. The panoramic lens of the binocular PTZ camera is activated, and the lens is simultaneously controlled to move horizontally left and right. The preset field of view of the panoramic lens is set to 120° to fit the basketball training hall's area. The movement range is set from the lens angle corresponding to the left sideline of the training hall to the lens angle corresponding to the right sideline (total movement angle is 90°), and the movement speed is set to 8° / s. Simultaneously, the panoramic lens's acquisition resolution is set to 1920×1080 pixels (balancing clarity and data processing speed). During the left and right movement, the lens continuously acquires images at a frequency of 2 frames per second, covering the entire training area. Image data is transmitted to the system terminal in real time, achieving comprehensive capture of athletes and basketballs within the training area and completing target detection.

[0027] This embodiment performs target detection by activating and controlling the left and right movement of a panoramic lens in a sports scene. By leveraging the wide field of view of the panoramic lens and the scene-adaptive movement and acquisition parameters, it achieves full-area coverage acquisition of the sports scene, solving the problems of limited field of view or target omission in existing single-lens detection. It provides a reliable preliminary data foundation for subsequent target recognition, localization, and close-up tracking, meeting the comprehensive target detection needs of scenarios such as live sports broadcasts and training data analysis.

[0028] Step S20: Identify the tracking target in the sports scene and simultaneously acquire the position information of the tracking target within the field of view of the panoramic lens; Optionally, the tracking target refers to the object in the sports scene that needs to be continuously monitored and tracked, including but not limited to the ball used in the competition, athletes, and other subjects with dynamic movement characteristics; the field of view of the panoramic lens refers to the spatial area of ​​the image that the panoramic lens can capture, which is determined by the field of view angle of the lens; the position information refers to the orientation data of the tracking target within the field of view of the panoramic lens, used to characterize the spatial position of the target in the scene; the global image is an image that covers the entire sports scene and is captured by the panoramic lens, containing visual information of all potential targets in the scene.

[0029] Optionally, after the panoramic lens completes the acquisition of the entire football field image, image feature analysis is performed on the acquired image. By distinguishing the target's outline, motion state, and color features, the football and players in the image are identified, irrelevant personnel and equipment on the sidelines are excluded, and the football is determined as the core tracking target. Simultaneously, when acquiring position information, a two-dimensional coordinate system is established with the upper left corner of the panoramic lens's field of view as the origin. Based on the pixel position of the football in the entire image, the coordinate data of the football in this coordinate system is calculated. This coordinate data represents the position information of the football within the panoramic lens's field of view, thus fully realizing the synchronous execution of target recognition and position acquisition.

[0030] Optionally, after the panoramic lens acquires a full-area image of the basketball training hall, it distinguishes between training athletes and sparring partners based on the target's movement trajectory and limb characteristics, and identifies the athlete whose movements need to be analyzed as the tracking target. When acquiring position information, it combines the focal length parameters and image pixel density of the panoramic lens to convert the athlete's pixel coordinates in the full-area image into relative position data within the actual training field. This data is the athlete's position information within the panoramic lens's field of view, ensuring accurate correspondence between the position information and the actual scene.

[0031] This embodiment solves the problems of target recognition confusion and insufficient position acquisition accuracy in the prior art by realizing the simultaneous recognition of tracking targets and acquisition of position information in sports scenarios. It provides accurate target objects and positional basis for subsequent control of close-up shots to rotate to the target position and start real-time tracking, thus meeting the needs of sports event live broadcasts for accurate target positioning and training data analysis for reliable position data.

[0032] Step S30: Based on the location information, control the close-up camera to rotate to the location of the target being tracked, and start the close-up camera to track the target in real time.

[0033] Optionally, close-up shots: lenses mounted on binocular pan-tilt cameras, with high-definition imaging and zoom capabilities; tracking target location refers to the specific direction and position of the tracked target in physical space, calculated based on the installation position of the binocular pan-tilt camera and combined with position information; real-time tracking refers to the close-up shot continuously following the movement of the tracked target, dynamically adjusting the shooting angle and focal length.

[0034] Optionally, this embodiment is applied to a live football match scenario. Currently, in sports events, existing close-up control largely relies on manual operation. Operators manually adjust the lens direction by viewing the panoramic image, resulting in a response delay of at least 1 to 2 seconds, which can easily cause them to miss fast-moving footballs or key player actions. In this embodiment, after acquiring the position information of the football (two-dimensional coordinate data with the binocular PTZ camera as the origin), the position information is transmitted to the camera's control module. The control module calculates the horizontal and vertical angles required for the close-up shot to rotate based on the coordinate data—for example, when the position information is (X1, Y1), it is calculated that the horizontal rotation is 30° and the vertical rotation is 5°. The control module generates a drive signal to drive the PTZ structure of the binocular PTZ camera to rotate, causing the close-up shot to turn towards the location of the football. Once the close-up shot is aligned with the football, the lens's shooting function is activated, continuously shooting at 1080P resolution and a frame rate of 30 frames per second. The position of the football in the frame is detected in real time through an image recognition algorithm, and the lens angle is dynamically fine-tuned to ensure that the football is always in the center of the frame, achieving real-time tracking.

[0035] Optionally, this embodiment is applied to a basketball player training data analysis scenario. In existing basketball training, close-up tracking often uses a fixed trajectory preset method, which cannot adapt to the flexible and ever-changing movement routes of athletes, resulting in tracking interruption or target deviation from the frame. In this embodiment, after obtaining the position information of the training athlete (including the relative distance and angle data of the athlete in the training hall), the position information is input into the system's motion prediction unit. This unit predicts the short-term movement direction of the athlete based on the athlete's current position and historical movement trends. The control module combines the real-time position and the predicted direction to calculate the rotation parameters of the close-up camera, driving the gimbal to quickly turn the close-up camera to the athlete's location. During the rotation, the position information is updated in real time, and the rotation angle is dynamically corrected to avoid lag. Furthermore, after the close-up camera is aimed at the athlete, a real-time tracking mode is activated, acquiring images at 720P resolution and a frame rate of 60 frames per second, focusing on capturing the details of the athlete's limb movements, and simultaneously transmitting the image data to the analysis terminal, providing clear and continuous visual data for training movement evaluation.

[0036] This embodiment achieves automatic turning and real-time tracking of close-up shots based on accurate position information, solving the problems of high latency in manual control and poor flexibility of fixed preset tracking in the prior art; it can quickly lock onto the tracking target without manual intervention, and the tracking process can dynamically adapt to the target's motion state, ensuring that the target details are clear and remain in the picture, meeting the real-time and viewing requirements of live sports events, as well as the requirements of motion capture accuracy for training data analysis.

[0037] This embodiment activates and controls the panoramic lens to move left and right to detect targets in the sports scene, achieving full-area coverage monitoring of the sports scene; it identifies the tracking target in the sports scene and simultaneously acquires the position information of the tracking target within the field of view of the panoramic lens, achieving accurate positioning of key tracking targets; based on the position information, it controls the close-up lens to rotate to the location of the tracking target and activates the close-up lens to track the tracking target in real time, achieving high-definition and real-time capture of the tracking target's motion state and action details, thereby improving the stability of target tracking in sports scenes.

[0038] Furthermore, based on the above, a second embodiment of the target tracking method of this embodiment is proposed. In some feasible embodiments, the system includes a target detection model, and the above step S10 includes the following implementation steps B201-B202.

[0039] Step B201: Start and control the panoramic lens to move left and right to acquire a full-field image of the sports scene at a preset first resolution, wherein the panoramic lens is a telephoto lens with a preset field of view. Step B202: Call the object detection model to perform AI detection on the entire image to identify the tracking target in the sports scene.

[0040] Optionally, the target detection model refers to an algorithm model used to identify and classify targets in an image. In this embodiment, image features are analyzed using, but not limited to, AI algorithms to achieve accurate target identification in sports scene tracking. This model can be pre-trained and deployed in the target tracking system. The preset first resolution refers to the pre-configured panoramic lens image acquisition resolution, with values ​​ranging from, but not limited to, 1920×1080 pixels to 3840×2160 pixels, which can be adapted to the scale of the sports scene. The preset field of view refers to the pre-set angle range of the panoramic lens that can acquire images, determining the size of the lens coverage area. In this embodiment, the value range includes, but is not limited to, 120°-180°. To meet the full-area coverage needs of different sports scenarios; a telephoto lens refers to a lens with a long focal length, which has the characteristics of clear imaging at long distances and shallow depth of field. In this embodiment, it is mounted on, but is not limited to, a binocular pan-tilt camera, which can clearly capture full-area images of sports scenarios at long distances; a full-area image refers to an image captured by a panoramic lens that completely covers the entire sports activity area, containing visual information of all potential tracking targets in the scene, and is the basic data for AI detection by the target detection model; AI detection refers to the process of feature extraction, target matching and classification of the captured full-area image through the artificial intelligence algorithm of the target detection model, which can automatically identify tracking targets in the image and distinguish target types.

[0041] Optionally, this embodiment is applied to a live football match scenario. The panoramic lens of the binocular PTZ camera (a telephoto lens) is activated, and control signals are sent to the PTZ drive module to control the lens to move horizontally left and right. The preset field of view is set to 150°, the preset first resolution to 1920×1080 pixels, and the movement range is the lens angle corresponding to the left and right sidelines of the field (total movement angle 120°). The movement speed is 10° / s, and the lens pauses for 0.5 seconds to capture images every 10° of movement. The panoramic lens continuously captures full-field images while moving left and right, covering the entire field from one goal to the other. The captured images are transmitted in real-time to the system processing unit, which calls the target detection model for AI detection. The model extracts the shape contour and color features of targets in the images (such as the black and white blocks of the football and the colors of the players' jerseys), automatically identifies the football and players on the field, excludes irrelevant objects on the sidelines, and locks onto the target for tracking.

[0042] Optionally, this embodiment is applied to a tennis match live broadcast scenario. Existing tennis match image acquisition often suffers from the small court size but rapid movement of the target (tennis ball). Using a short-focal-length lens results in the target appearing too small, while using a telephoto lens for fixed shooting limits the field of view. Furthermore, general models lack sufficient accuracy in recognizing small targets like tennis balls. In this embodiment, a panoramic lens (telephoto lens) is activated, and its horizontal movement is controlled simultaneously. A preset field of view of 120° and a preset first resolution of 2560×1440 pixels are set. The movement range is the lens angle corresponding to the left and right baselines of the tennis court (total movement angle 90°), with a movement speed of 8° / s. The entire field of view is continuously acquired during the lens movement. After the image is transmitted to the processing unit, an AI detection model optimized for small targets is invoked for detection. The model, through an enhanced small target feature extraction module, accurately identifies the tennis ball and the athlete, clearly tracking the target.

[0043] This embodiment achieves high-quality image acquisition across the entire sports scene by activating and controlling the panoramic lens to move left and right to acquire images, combined with the characteristics of the telephoto lens and the parameter configuration adapted to the scene; by leveraging the AI ​​detection capabilities of the target detection model, it automatically and accurately identifies and tracks targets, solving the problems of imbalance between lens field of view and sharpness, and the reliance on manual or low accuracy of target recognition in existing technologies. It provides high-quality data and accurate target objects for subsequent acquisition of target positions and initiation of close-up tracking, meeting the needs of live sports broadcasts for full coverage and accurate target identification.

[0044] Furthermore, based on the above, in some feasible embodiments, the preset verification algorithm includes a unit root test algorithm and a periodic feature detection algorithm, and the above step S20 includes the following implementation steps C301-C303.

[0045] Step C301: Sort multiple candidate tracking targets in the sports scene identified by target detection according to a preset priority rule; Step C302: Select the tracking target with the highest priority after sorting, and use the tracking target with the highest priority as the updated tracking target; Step C303: Based on the updated pixel coordinates and pixel percentage information of the tracked target in the global image, simultaneously acquire the updated spatial position information of the tracked target within the field of view of the panoramic lens, and use the spatial position information as the position information.

[0046] Optionally, candidate tracking targets refer to objects that may need to be tracked, identified by the target detection model after detecting the entire sports scene image, including but not limited to athletes, event balls, referees, and other subjects with dynamic characteristics; preset priority rules refer to pre-configured criteria for prioritizing candidate tracking targets, including but not limited to sorting standards based on target type, scene requirements, motion state, etc., which can be flexibly adjusted according to different sports scenes; updated tracking targets refer to the candidate tracking targets with the highest priority selected after priority sorting; pixel coordinates refer to the position coordinates of the tracking target in the entire image, in units of image pixels; pixel proportion information refers to the ratio of the number of pixels occupied by the tracking target in the entire image to the total number of pixels in the image; spatial position information refers to the position data of the tracking target within the field of view of the panoramic lens, including but not limited to the horizontal and vertical angle coordinates based on the installation position of the panoramic lens, or two-dimensional plane coordinates established based on the field of view.

[0047] Optionally, in current football matches, existing technologies often handle multiple candidate tracking targets through random selection or manual designation. Random selection is prone to missing core targets (such as the football), while manual designation suffers from delays and cannot adapt to the rapidly changing situation on the field. In this embodiment, after the target detection model detects the entire football field image, the identified candidate tracking targets include, but are not limited to, the football, 22 players, and 3 referees. The system's built-in preset priority rule is set based on live broadcast requirements as follows: football > main players on the field > ordinary players > referees. The processing unit sorts the candidate tracking targets according to this rule, and after sorting, selects the football, which has the highest priority, as the updated tracking target. Subsequently, the pixel coordinates of the soccer ball in the global image (e.g., (800, 600, with the top left corner of the image as the origin) and pixel percentage information (e.g., percentage of 0.5%) are obtained. Combined with the focal length parameters of the panoramic lens and the image pixel density (in this embodiment, the pixel density is 10 pixels / meter), the pixel coordinates are converted into spatial position information within the field of view of the panoramic lens. For example, with the bottom left corner of the field of view of the panoramic lens as the origin, the two-dimensional coordinates of the soccer ball are calculated to be (40 meters, 30 meters). These coordinates are the position information for subsequent control of close-up shots.

[0048] Optionally, in basketball player training data analysis scenarios, existing basketball training often overlooks training priorities when selecting candidate tracking targets, mistakenly identifying sparring partners as the core tracking object, leading to biased training data collection. In this embodiment, after the target detection model detects the entire training hall image, the identified candidate tracking targets include, but are not limited to, the target training athlete, the basketball, and three sparring partners. The preset priority rule is set based on training requirements as: target training athlete > basketball > sparring partners. After sorting according to this rule, the processing unit selects the target training athlete as the updated tracking target. The pixel coordinates (e.g., (1200, 900)) and pixel percentage information (e.g., percentage 3%) of the athlete in the entire image are obtained. Combined with the preset field of view (120°) and image resolution (1920×1080 pixels) of the panoramic lens, the spatial position information of the athlete within the field of view of the panoramic lens (e.g., horizontal angle 25°, vertical angle 10°) is calculated and transmitted as position information to the close-up control module.

[0049] This embodiment ensures that no core tracking target is missed by using preset priority rules adapted to the scene. Combined with the conversion logic of pixel coordinates and pixel ratio, it ensures the accuracy of position information and provides a reliable target object and position basis for the accurate turning and real-time tracking of subsequent close-up shots. This meets the needs of sports event live broadcasts for focusing on core targets and training data analysis for accurate target positioning.

[0050] Furthermore, based on the above, in some feasible embodiments, step S30 above may also include the following implementation steps D401 to D403.

[0051] Step D401: Based on the position information, generate drive commands for controlling the rotation of the close-up shot; Step D402: Based on the drive command, control the stepper motor to drive the close-up camera to rotate to the location of the tracking target; Step D403: Start the close-up shot to track the target in real time at a preset frame rate and a preset second resolution.

[0052] Optionally, the drive command refers to the control signal generated based on the position information of the tracked target, used to control the rotation of the close-up lens, including but not limited to the command data for controlling parameters such as the rotation direction, rotation angle, and rotation speed of the lens; the stepper motor refers to an actuator that converts electrical pulse signals into angular or linear displacement, which in this embodiment is mounted on a binocular pan-tilt camera to receive drive commands and drive the close-up lens to rotate precisely; the preset frame rate refers to the pre-configured image acquisition frequency for real-time tracking of the close-up lens, measured in frames per second (fps), with a value range including but not limited to 25fps-60fps, which can be adjusted according to the movement speed of the tracked target; the preset second resolution refers to the pre-configured image acquisition resolution for real-time tracking of the close-up lens, with a value range including but not limited to 1280×720 pixels-1920×1080 pixels, used to balance the detail clarity of the tracking image with data processing efficiency.

[0053] Optionally, this embodiment is applied to a live football match broadcast scenario. Currently, in football matches, existing close-up shot rotation control mostly relies on manual operation. Operators manually adjust the lens after judging the target's location based on the panoramic view, resulting in a 1-2 second response delay, which can easily cause them to miss crucial moments such as shots or player breakthroughs. Some solutions using motor control suffer from insufficient lens rotation precision due to the coarse basis of the drive command generation, making it difficult to accurately align with the fast-moving football. In this embodiment, after obtaining the position information of the football within the panoramic lens's field of view (e.g., a horizontal angle of 35° and a vertical angle of 8°), the system's control module converts this position information into lens rotation parameters—using the initial zero position of the binocular pan-tilt camera as a reference, it calculates that the close-up shot needs to rotate 35° horizontally clockwise and 8° vertically upward, while setting the rotation speed to 5° / ms. Based on these parameters, a drive command is generated. After this drive command is transmitted to the stepper motor's drive module, the stepper motor is controlled to operate according to the command parameters. The stepper motor drives the close-up shot to rotate through a gear transmission structure. During the rotation, the rotation angle is fed back to the control module in real time for deviation correction, ensuring that the close-up shot is accurately rotated to the location of the football. Once the camera is focused on the football, the real-time tracking mode is activated, with a preset frame rate of 30fps and a preset second resolution of 1920×1080 pixels. The close-up camera continuously captures the movement of the football using these parameters, ensuring that the details of the football are clear and its movement trajectory is smooth in the live broadcast.

[0054] Optionally, this embodiment is applied to the data analysis scenario of track and field sprint training. In existing track and field training, close-up tracking often results in stuttering due to the high speed of the target (the instantaneous speed of sprinters can reach over 10m / s) and a low frame rate configuration, or the camera cannot follow the target's movement in time due to motor response lag, affecting the accuracy of motion data acquisition. In this embodiment, after obtaining the sprinter's position information (e.g., horizontal angle 15°, vertical angle 5°), the control module combines the athlete's current movement speed (8m / s) to generate an appropriate drive command—setting the rotation speed to 8° / ms to ensure that the camera can quickly keep up with the athlete's movement rhythm, while also including rotation angle parameters (horizontal clockwise 15°, vertical upward 5°). After the drive command is sent to the stepper motor, the motor is controlled to drive the close-up camera to rotate quickly and accurately to the athlete's location. Then, the close-up camera was activated for real-time tracking. Considering the high speed of the athlete's movement, the preset frame rate was set to 60fps to ensure smooth footage, and the preset second resolution was 1280×720 pixels to reduce data processing pressure. The close-up camera used these parameters to capture details of the athlete's starting, acceleration, arm swing, and other movements, and simultaneously transmitted the image data to the training analysis terminal to provide clear and continuous data support for movement optimization.

[0055] This embodiment generates precise drive commands based on location information, and combines the high-precision rotation of the stepper motor to ensure that the lens quickly aligns with the target; the preset frame rate and preset second resolution adapted to the scene balance the smoothness of the picture and the clarity of details, providing reliable technical support for capturing key moments in live sports events and accurately collecting motion data for training data analysis, and meeting the needs of target tracking accuracy and real-time performance in different sports scenarios.

[0056] Furthermore, based on the content of any of the above embodiments, in some feasible embodiments, step D403 may further include the following implementation steps E501-E504.

[0057] Step E501: Start the close-up shot. When the close-up shot fails to detect the tracked target for a preset duration at a preset frame rate and preset second resolution, trigger the target tracking switching mechanism. Step E502: When the close-up lens switches from tracking the target to tracking the target in the panoramic lens, stop the active tracking action of the close-up lens and obtain the motion trajectory of the close-up lens before losing tracking of the target; Step E503: Based on the motion trajectory, determine the extended movement direction of the panoramic lens and control the panoramic lens to move in the extended movement direction. The extended movement direction is used to cover the blind spot in the field of view when the close-up shot loses tracking of the target. Step E504: When the panoramic lens is detected to move to the extended movement direction, control the panoramic lens to perform global target detection on the sports scene at a resolution greater than or equal to a preset first resolution, so as to continuously identify the tracking targets in the sports scene.

[0058] Optionally, the duration of continuous unidentified tracking target refers to the length of time during which the close-up lens fails to capture an image of the tracking target within the lens's field of view during real-time tracking at a preset frame rate and preset second resolution; the target tracking switching mechanism refers to the control logic triggered when the close-up lens cannot identify the tracking target, switching from "close-up tracking mode" to "panoramic lens full-field detection mode," including but not limited to processes such as generating mode switching instructions and adjusting lens working status; active tracking action refers to the continuous action of the close-up lens to actively adjust the shooting angle, focal length, etc., to keep the tracking target in the center of the field of view, including but not limited to operations such as lens rotation and zoom adjustment; motion trajectory refers to the information on the movement path of the tracking target in space recorded by continuously acquired image data before the close-up lens loses the tracking target, including but not limited to the target's position coordinates and movement direction at different times; extended movement direction refers to the direction that the panoramic lens needs to move based on the predicted motion trajectory before the tracking target is lost; the blind spot of the field of view refers to the area where the close-up lens cannot acquire images due to the limitation of the field of view or obstruction by obstacles at the current tracking position.

[0059] Optionally, this embodiment is applied to a live football match scenario. A close-up camera is activated to track the football in real time. When the close-up camera fails to detect the football continuously at 30fps and 1920×1080 pixel resolution for a preset duration (set to 1 second in this embodiment), the system triggers a target tracking switching mechanism. Upon detecting the switching mechanism, the close-up camera immediately stops its active tracking action, ceases angle adjustments and zooming, and retrieves the football's trajectory recorded in the 500 milliseconds prior to its loss (including three sets of position coordinates showing the football moving from the left side of the penalty area towards the goal). Based on this trajectory, it is predicted that the football may continue moving towards the goal, and the extended movement direction of the panoramic camera is determined to be towards the goal (horizontally rotated 20° to the right). The panoramic camera is then controlled to move in this direction. Once the panoramic camera completes its rotation, it is controlled to perform full-area target detection across the field, including the goal area, with parameters no lower than a preset first resolution (1920×1080 pixels), continuously capturing the football's dynamics.

[0060] Optionally, this embodiment is applied to a basketball player training scenario. The close-up camera tracks the training athlete at a frame rate of 60fps and a resolution of 1280×720 pixels. When the duration of continuous non-identification of the athlete reaches a preset duration (0.8 seconds in this embodiment), a target tracking switching mechanism is triggered. The system stops the active tracking action of the close-up camera and obtains the athlete's movement trajectory before being lost (recording the movement path of the athlete sprinting from the three-point line to the basket). Based on the trajectory prediction, it is determined that the athlete may continue to move towards the basket, and the direction of the panoramic lens expansion movement is determined to be towards the basket (horizontally rotating downwards by 15°). The panoramic lens is then controlled to move in this direction. After the panoramic lens is in position, a full-area detection is performed on the area under the basket and surrounding areas with parameters greater than the preset first resolution (set to 2560×1440 pixels) to quickly locate the athlete's position.

[0061] In this embodiment, when the close-up shot loses tracking of the target, the detection range of the panoramic lens can be precisely adjusted based on the target's motion trajectory, avoiding blind scanning by the panoramic lens and significantly shortening the time for re-identifying the target. At the same time, by promptly stopping the close-up shot's active tracking and controlling the directional movement of the panoramic lens, the smoothness of tracking switching is ensured. This solves the problems of low re-identification efficiency and missing key scenes after target loss in the prior art, meeting the real-time and continuous requirements of live sports broadcasts and the continuous target tracking requirements of training scenarios.

[0062] Furthermore, based on the content of any of the above embodiments, in some feasible embodiments, the target tracking method further includes the following implementation steps F501 to F503.

[0063] Step F501: Collect the motion parameters of the tracked target in real time, and calculate the moving speed of the tracked target based on the motion parameters. The motion parameters include the displacement distance and direction of movement of the tracked target per unit time. Step F501: Collect historical occlusion data of the tracked target, and calculate the occlusion probability of the tracked target by combining it with the current obstacle distribution in the scene; Step F501: Dynamically adjust the preset duration based on the relationship between movement speed and occlusion probability.

[0064] Optionally, motion parameters refer to data used to characterize the motion state of the tracked target, including but not limited to the displacement distance and direction of motion of the tracked target per unit time; displacement distance refers to the spatial span of the tracked target moving along the direction of motion within a set unit time, with units including but not limited to meters (m) and centimeters (cm); direction of motion refers to the direction of movement of the tracked target relative to the panoramic lens's field of view coordinate system during motion, including but not limited to horizontal angles, vertical angles, etc.; historical occlusion data refers to relevant information recorded by the system before the current tracking process regarding when the tracked target was occluded by obstacles in the scene, including but not limited to the time point of occlusion, the duration of a single occlusion, and the frequency of occlusion per unit time; the current distribution of obstacles in the scene refers to the situation in the sports scene where the tracked target is located, which may affect the imaging of the target. The real-time status of occluded objects includes, but is not limited to, the position, quantity, and relative distribution of other objects such as athletes, equipment, and referees. Occlusion probability refers to the probability that the target will be occluded in the short term based on the historical occlusion patterns of the tracked target and the current obstacle distribution in the scene, expressed as a percentage (%), with a value range including but not limited to 0% to 100%. Correlation refers to the corresponding rules that affect the preset duration adjustment by the tracked target's movement speed and occlusion probability, including but not limited to the weighting of the two on the preset duration and the calculation logic of the numerical adjustment range. Dynamically adjusting the preset duration means updating the specific value of the preset duration in real time according to the preset correlation based on the real-time acquired movement speed and occlusion probability, so that the preset duration adapts to the real-time movement status of the tracked target and changes in the scene environment.

[0065] Optionally, this embodiment is applied to a live football match scenario. Currently, in football matches, existing technologies often use a fixed value for the preset duration (e.g., uniformly set to 2 seconds), without considering changes in the football's movement speed and scene occlusion. When the football is moving at high speed (e.g., at the moment of a shot), a fixed duration can easily lead to a delay in switching after tracking is lost; when scene occlusion is frequent (e.g., when players are densely packed in the penalty area), it is easy to accidentally trigger a switch due to brief occlusion. In this embodiment, the system collects the football's motion parameters in real time: the unit time is set to 0.5 seconds, and the image data continuously collected through close-up shots is used to extract the football's displacement distance (e.g., 5 meters) and direction of movement (towards the left side of the goal) within 0.5 seconds. Based on the formula "movement speed = displacement distance / unit time", the football's movement speed is calculated to be 10 m / s. Simultaneously, historical occlusion data of the football was analyzed: tracking records from the past 10 minutes were retrieved, showing that the football was occluded 3 times by players, with an average duration of 0.3 seconds per occlusion and an occlusion frequency of 0.3 times per minute. Combined with the current obstacle distribution in the scene (3 defending players and 1 attacking player in the penalty area, forming a densely populated area around the football), the occlusion probability was calculated as 35% using the formula: "Occlusion probability = (Historical occlusion frequency × Current obstacle density coefficient) × 100%". Based on preset correlations: movement speed is negatively correlated with preset duration (for every 2m / s increase in speed, duration decreases by 0.1 seconds), while occlusion probability is positively correlated with preset duration (for every 10% increase in probability, duration increases by 0.2 seconds). Using a base duration of 1 second as a baseline, the adjusted preset duration was calculated to be 1.1 seconds, thus completing the dynamic update of the preset duration.

[0066] Optionally, this embodiment is applied to a basketball player training scenario. In existing basketball training, the preset duration is often set according to the athlete's average speed (e.g., 1.5 seconds). When the athlete suddenly changes speed (e.g., switching from jogging to sprinting) or when obstacles in the scene change (e.g., a training partner approaches), the fixed duration cannot adapt, easily leading to untimely tracking or false triggering. In this embodiment, the athlete's motion parameters are collected in real time: the unit time is set to 0.3 seconds, and the athlete's displacement distance within 0.3 seconds is captured by a close-up shot as 2.1 meters, with the direction of movement towards the basket, and the calculated movement speed is 7 m / s. The athlete's historical occlusion data was analyzed: within the past 5 minutes, the athlete was obscured twice by a training partner, with each occlusion lasting 0.2 seconds, resulting in an occlusion frequency of 0.4 times per minute. Considering the current obstacle distribution in the scene (e.g., a training partner is approximately 1.5 meters away from the athlete in the area under the basket, posing a potential occlusion risk), the occlusion probability was calculated to be 25%. Based on the correlation: a movement speed of 7 m / s corresponds to a base duration of 0.7 seconds, and a 25% occlusion probability corresponds to an additional 0.5 seconds of adjustment. The final dynamically adjusted preset duration is 1.2 seconds, adapting to the athlete's current movement state and the scene environment.

[0067] This embodiment calculates speed by collecting motion parameters in real time, combines historical and current data to statistically analyze occlusion probability, and then adjusts the duration based on the correlation to ensure that the preset duration always matches the target motion and scene occlusion. This effectively balances the sensitivity of tracking switching and the ability to resist false triggering, ensuring the stability and accuracy of live sports broadcasts and training tracking.

[0068] Furthermore, based on the content of any of the above embodiments, in some feasible embodiments, the target tracking method described above, after step E504, further includes steps G40-G50.

[0069] Step G40: When the panoramic lens re-identifies the tracked target, update the position information of the tracked target; Step G50: Based on the updated location information of the tracked target, return to the step of controlling the close-up camera to rotate to the location of the tracked target based on the location information, and start the close-up camera to track the target in real time.

[0070] Optionally, re-identifying the tracking target refers to the process by which the panoramic lens, during the full-domain detection process, recaptures the previously lost tracking target image through image analysis and confirms that the image matches the tracking target features before it was lost. This is a prerequisite for triggering subsequent position information updates and tracking switching. Updating the tracking target's position information refers to recalculating and replacing the original tracking target position data based on the image data collected when the panoramic lens re-identifies the tracking target, ensuring that the position information is consistent with the target's current actual spatial orientation. This includes, but is not limited to, updating the target's pixel coordinates and two-dimensional coordinates within the field of view. Returning to execution refers to restarting the previously executed process of "controlling close-up rotation and real-time tracking based on position information" after completing the update of the tracking target's position information, so that the target tracking switches from the "panoramic full-domain detection mode" back to the "close-up precise tracking mode" control logic.

[0071] Optionally, this embodiment is applied to a live football match scenario. Currently, in football matches, existing technologies, after the panoramic lens re-identifies the football, often directly use the position information before it was lost to drive the close-up camera to rotate, or require manual confirmation before tracking can be started. The former is prone to lens aiming errors due to positional deviations, while the latter has a 1-2 second delay, missing crucial footage of the football's subsequent movement. In this embodiment, when the panoramic lens performs full-domain detection in the extended movement area in the direction of the goal, it uses a target detection model to capture an image that matches the characteristics of the football before it was lost (black and white block texture, circular outline), thus determining that the tracking target has been re-identified. Then, based on the pixel coordinates of the soccer ball in the image (e.g., (950, 550, with the top left corner of the global image as the origin) and the current field of view parameters of the panoramic lens (field of view 150°, resolution 1920×1080 pixels), the spatial position information of the soccer ball within the field of view of the panoramic lens (horizontal angle 38°, vertical angle 7°) is calculated, and the position information is updated. Based on the updated position information, the system generates a drive command to control the stepper motor to drive the close-up lens to rotate to that position. After the lens is aligned with the soccer ball, the close-up lens is started to track the soccer ball in real time at a frame rate of 30fps and a resolution of 1920×1080 pixels, restoring the accurate tracking state.

[0072] Optionally, this embodiment is applied to basketball player training scenarios. In existing basketball training, after the panoramic camera re-identifies the athlete, the position information update often ignores the athlete's real-time movement trend. This results in the athlete having moved again by the time the close-up shot is in position, requiring a second adjustment and affecting the continuity of training data acquisition. In this embodiment, when the panoramic camera detects the direction of expansion under the basket, it identifies a target that matches the characteristics of the athlete before the loss of training (specific training uniform color, limb outline), confirming successful re-identification. Based on the athlete's pixel coordinates in the identified image (e.g., (1100, 600)) and the panoramic camera's field of view parameters (field of view angle 120°, resolution 2560×1440 pixels), the athlete's current spatial position information (horizontal angle 22°, vertical angle 9°) is calculated. At the same time, combined with the athlete's limb posture (leaning forward, stepping motion) in the image, the direction of movement is predicted, and the position information is dynamically corrected to complete the update. Based on the corrected position information, a drive command containing a predicted compensation angle is generated to control the stepper motor to drive the close-up lens to rotate. During the lens rotation, minute updates of position information are received synchronously, and the rotation angle is adjusted in real time to ensure accurate alignment with the athlete. Then, the close-up lens is started to perform real-time tracking at a frame rate of 60fps and a resolution of 1280×720 pixels, continuously collecting the athlete's motion data.

[0073] Optionally, such as Figure 3 As shown, Figure 3The flowchart of the target tracking method involved in this application is as follows: The process starts with "start tracking", first enters the "close-up and panoramic tracking start position" step to complete the initial position calibration of the dual lenses of the binocular pan-tilt camera; then, "panoramic lens detects target" is executed, which uses the wide field of view of the panoramic lens to detect the entire sports scene and initially identify potential tracking targets; after obtaining the target position, the process enters the "rotate the close-up to the specified position and start detection on the close-up" step, which controls the close-up to accurately turn to the target and start high-definition detection based on the position data. Then, the panoramic lens is moved in reverse to ensure the overall detection range. The panoramic lens is moved in reverse to expand the system's detection coverage area. When the target is lost in the close-up shot, the system switches to the panoramic lens for detection. If the target is lost in the close-up shot, the system immediately switches to the panoramic lens for full-area detection. The system then enters the judgment step "Detection timeout 10s". If the timeout occurs, the system returns to the "Close-up and panoramic tracking start position" for re-initialization. If the timeout does not occur, the system returns to the "Rotate the close-up shot to the specified position and start detection on the close-up shot" step to continue tracking. The entire process achieves continuous and accurate target tracking through dual-lens collaborative switching, position calibration, and timeout mechanisms, adapting to the complex environment of dynamic target movement in sports scenarios.

[0074] This embodiment ensures that close-up shots can quickly and accurately target and resume tracking by updating location information in real time and combining scene-adaptive tracking parameters. This guarantees the continuity of live sports broadcasts and the integrity of training data collection, meeting the needs for continuous and accurate target tracking in different sports scenarios.

[0075] In addition, this application also provides a target tracking system, please refer to... Figure 4 , Figure 4 This is a schematic diagram of the target tracking system involved in the embodiments of this application. The target tracking system provided in this application includes: The target detection module H01 is used to start and control the left and right movement of the panoramic camera to perform target detection in the sports scene. The information acquisition module H02 is used to identify the tracking target in the sports scene and simultaneously acquire the position information of the tracking target within the field of view of the panoramic lens; The lens control module H03 is used to control the close-up lens to rotate to the location of the target based on the position information, and to start the close-up lens to track the target in real time.

[0076] The target tracking system provided in this application, employing the target tracking method described in the above embodiments, can solve the technical problem of insufficient target tracking stability in target tracking systems. Compared with the prior art, the beneficial effects of the target tracking system provided in this application are the same as those of the target tracking method described in the above embodiments, and other technical features of this target tracking system are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0077] In addition, this application also provides an electronic device. Please refer to... Figure 5 , Figure 5 This is a schematic diagram of the structure of the electronic device involved in the embodiments of this application.

[0078] This application provides an electronic device, which includes: at least one processor; a binocular pan-tilt camera; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the target tracking method in Embodiment 1 above.

[0079] The following is for reference. Figure 5 , Figure 5 This is a schematic diagram of the structure of an electronic device involved in the embodiments of this application, illustrating a structural schematic diagram of an electronic device suitable for implementing the embodiments of this application. The electronic devices in the embodiments of this application may include, but are not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), in-vehicle terminals (such as in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 5 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0080] like Figure 5As shown, the electronic device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the electronic device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following devices may be connected to the I / O interface 1006: input devices 1007 including, for example, a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. Communication device 1009 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although electronic devices with various devices are shown in the figures, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively.

[0081] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0082] The electronic device provided in this application, employing the target tracking method described in the above embodiments, can solve the technical problem of insufficient target tracking stability in sports scenarios. Compared with the prior art, the beneficial effects of the electronic device provided in this application are the same as those of the target tracking method provided in the above embodiments, and other technical features of this electronic device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0083] Furthermore, this application provides a computer-readable storage medium. This computer-readable storage medium stores a target tracking program, which, when executed by a processor, implements the steps of the target tracking method described above.

[0084] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0085] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0086] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a computer-readable storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0087] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A target tracking method characterized by, The application is applied to a target tracking system, the target tracking system comprises a binocular pan-tilt camera, the binocular pan-tilt camera comprises a panoramic lens and a close-up lens, and the target tracking method comprises the following steps: starting and controlling the panoramic lens to move left and right to detect a target in a sports scene; identifying a tracking target in the sports scene and synchronously acquiring position information of the tracking target in a field of view range of the panoramic lens; based on the position information, controlling the close-up lens to rotate to a direction of the tracking target and starting the close-up lens to track the tracking target in real time.

2. The object tracking method of claim 1, wherein, The target tracking system comprises a target detection model, and the step of starting and controlling the panoramic lens to move left and right to detect a target in a sports scene comprises the following steps: starting and controlling the panoramic lens to move left and right to collect a global image of the sports scene at a preset first resolution, wherein the panoramic lens is a long-focus lens with a preset field of view angle; calling the target detection model to perform AI detection on the global image to identify a tracking target in the sports scene.

3. The object tracking method of claim 2, wherein, The step of identifying a tracking target in the sports scene and synchronously acquiring position information of the tracking target in a field of view range of the panoramic lens comprises the following steps: sorting a plurality of candidate tracking targets in the sports scene identified by the target detection according to a preset priority rule; selecting a tracking target with the highest priority after sorting as an updated tracking target; based on pixel coordinate and pixel proportion information of the updated tracking target in the global image, synchronously acquiring spatial position information of the updated tracking target in the field of view range of the panoramic lens, and taking the spatial position information as the position information.

4. The object tracking method of claim 3, wherein, The target tracking system comprises a step motor, and the step of based on the position information, controlling the close-up lens to rotate to a direction of the tracking target and starting the close-up lens to track the tracking target in real time comprises the following steps: based on the position information, generating a driving instruction for controlling the close-up lens to rotate; based on the driving instruction, controlling the step motor to drive the close-up lens to rotate to the direction of the tracking target; starting the close-up lens to track the tracking target in real time at a preset frame rate and a preset second resolution.

5. The object tracking method of claim 4, wherein, The step of starting the close-up lens to track the tracking target in real time at a preset frame rate and a preset second resolution further comprises the following steps: starting the close-up lens, and when a time length of continuously not identifying the tracking target by the close-up lens at the preset frame rate and the preset second resolution reaches a preset time length, triggering a target tracking switching mechanism; when the close-up lens tracking the tracking target switches to the panoramic lens to globally detect the tracking target, stopping the active tracking action of the close-up lens and acquiring a motion trajectory of the close-up lens before losing the tracking target. Based on the motion trajectory, a direction of an extension movement of the panoramic lens is determined, and the panoramic lens is controlled to move to the direction of the extension movement, wherein the direction of the extension movement is used to cover a field-of-view blind area when the close-up lens loses the tracking target; When it is detected that the panoramic lens moves to the direction of the extension movement, the panoramic lens is controlled to perform global target detection on the sports scene with a resolution greater than or equal to the preset first resolution, so as to continuously identify the tracking target in the sports scene.

6. The object tracking method of claim 5, wherein, The method further includes: Real-time acquisition of a motion parameter of the tracking target, and calculation of a moving speed of the tracking target based on the motion parameter, wherein the motion parameter includes a displacement distance and a motion direction of the tracking target in a unit time; Statistical calculation of historical occlusion data of the tracking target, and calculation of an occlusion probability of the tracking target in combination with a distribution of obstacles in a current scene; Dynamic adjustment of the preset time length according to a correlation between the moving speed and the occlusion probability.

7. The object tracking method of claim 5, wherein, After the step of performing global target detection on the sports scene, the method further includes: When the tracking target is re-identified by the panoramic lens, position information of the tracking target is updated; Based on the updated position information of the tracking target, the step of controlling the close-up lens to turn to a direction in which the tracking target is located and starting real-time tracking of the tracking target by the close-up lens is returned to be executed based on the position information.

8. A target tracking system characterized by, The target tracking system includes a binocular pan-tilt camera, and the binocular pan-tilt camera includes a panoramic lens and a close-up lens. A target detection module is configured to start and control the panoramic lens to move left and right, and perform target detection on a sports scene. An information acquisition module is configured to identify a tracking target in the sports scene, and synchronously acquire position information of the tracking target in a field-of-view range of the panoramic lens. A lens control module is configured to control the close-up lens to turn to a direction in which the tracking target is located based on the position information, and start real-time tracking of the tracking target by the close-up lens.

9. An electronic device, comprising: The electronic device includes a binocular pan-tilt camera, a processor, a memory, and a target tracking program stored in the memory and executable by the processor, and when the target tracking program is executed by the processor, steps of the target tracking method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a target tracking program, and when the target tracking program is executed by a processor, steps of the target tracking method according to any one of claims 1 to 7 are implemented.

Images