Intelligent shooting robot

Abstract

The invention discloses an intelligent shooting robot, and relates to the technical field of audio and video production. The recognition unit supports data input of various sensors, and the recognition unit can achieve high-precision target recognition, action recognition and posture estimation; the transmission unit is used for carrying out target tracking under the real-time or near-real-time condition, and is suitable for an application scene needing quick response; the calculation unit improves the processing speed and efficiency; the configuration unit is used for performing self-defined configuration according to different application scenes to meet specific requirements; the data analysis unit carries out training and optimization by using a large amount of historical data, so that the accuracy and robustness of the system are improved; the learning unit supports online learning and model updating, and the performance and the adaptive capacity of the system are continuously improved; all links from copywriting generation to video shooting, editing and publishing are automated, so that the production efficiency is greatly improved, manual operation errors are reduced, and the working efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of audio and video production, and in particular to an intelligent shooting robot. Background Art

[0002] Audio and video production refers to the process of collecting, editing, and synthesizing raw audio and video materials into works with specific themes, styles, and forms of expression. This process is widely used in the fields of film, TV series, advertising, documentaries, etc., and is an important part of the media industry. Audio and video production requires comprehensive consideration of multiple aspects such as art, technology, and business to achieve the best audio-visual effects and commercial value.

[0003] In today's booming short video shooting market, with the rapid growth of content creation demand, traditional manual shooting methods have exposed many limitations. Manual shooting not only consumes a lot of manpower and time costs, but also makes it difficult to ensure stable output in terms of camera accuracy, smoothness of multi-angle switching, and consistency of shooting effects. Some existing robotic arms also have the problem of insufficient functional adaptability when applied to the shooting field. For example, most robotic arms lack exclusive functional modules customized for the characteristics of short video shooting, and it is difficult to meet the fast, changeable, and creative shooting requirements of short videos. For this reason, we propose an intelligent shooting robot. Summary of the invention

[0004] The purpose of the present invention is to solve the problems mentioned in the above background technology, and the present invention provides an intelligent shooting robot.

[0005] In order to achieve the above-mentioned purpose, the present invention specifically adopts the following technical solutions:

[0006] An intelligent shooting robot, comprising:

[0007] Storage boxes;

[0008] An automatic shooting module is installed at the center of the upper end of the storage box, and the automatic shooting module includes a driving machine, a driving wheel is sleeved on the circumferential surface of the upper end of the driving shaft of the driving machine, a transmission wheel is transmission-connected to the outer side of the driving wheel, a hydraulic cylinder is installed on the upper end of the transmission wheel, a support seat is installed on the upper end of the hydraulic cylinder, and a camera is installed on the upper end of the support seat, and the camera has a built-in visual tracking module;

[0009] A display mechanism is installed on the outer wall of the left side of the storage box, and the display mechanism includes a motor, a moving column movably installed on the circumferential surface of the motor shaft of the motor, a supporting cylinder installed on the outer side of the moving column, a cylinder installed inside the supporting cylinder, and a display screen installed on the upper end of the cylinder;

[0010] A host computer is installed inside the storage box, and the host computer has an automatic editing module and a text generation module built in;

[0011] A switch, installed inside the storage box and located on the right side of the host, the switch having a built-in automatic publishing module;

[0012] A bottom plate, installed below the storage box;

[0013] Moving wheels, mounted at the four corners of the lower end of the base plate;

[0014] The cabinet door is detachably mounted on the front side of the storage box via a movable shaft.

[0015] Furthermore, a plurality of driving teeth are evenly mounted on the circumferential surface of the driving wheel, a plurality of driving teeth are evenly opened inside the lower end of the transmission wheel, and the driving wheel is meshingly connected with the transmission wheel.

[0016] Furthermore, the visual tracking module includes a recognition unit, a transmission unit, a computing unit, a configuration unit, a data analysis unit and a learning unit, wherein the recognition unit supports data input from multiple sensors, such as cameras, lidars, etc.; the transmission unit is used to track targets in real time or near real time, and is suitable for application scenarios that require fast response; the computing unit uses technologies such as GPU acceleration to improve processing speed and efficiency; the configuration unit is used to perform customized configuration according to different application scenarios to meet specific needs; the data analysis unit uses a large amount of historical data for training and optimization to improve the accuracy and robustness of the system; the learning unit supports online learning and model updates to continuously improve the performance and adaptability of the system.

[0017] Furthermore, the recognition unit includes target recognition, action recognition and posture estimation. The target recognition utilizes advanced deep learning algorithms, such as YOLO, FasterR-CNN, etc., to achieve high-precision target recognition; the action recognition is used to recognize and track various human actions, such as walking, running, jumping, etc.; the posture estimation analyzes the posture and motion trajectory of the human body through skeletal key point detection.

[0018] Furthermore, a movable groove with arc-shaped structures at both ends is opened inside the left side of the storage box, the right side of the movable column is movably connected with the storage box through the movable groove, and a threaded hole is opened inside the right side of the movable column, the circumferential surface of the motor shaft of the motor is provided with an external thread, and the movable column is connected to the motor shaft through threads.

[0019] Furthermore, the display screen uses a high-resolution display, and the display screen is equipped with a matching mouse and keyboard. A positioning frame is installed on the outer side of the display screen, and the display screen is installed in the positioning frame.

[0020] Furthermore, the automatic editing module is used to realize sorting, editing, adding subtitles, special effects, color adjustment and audio processing through intelligent processing of the background system after the user completes the shooting. The copy generation module uses the advanced AI large language model and the independently designed prompt word system, combined with the fine-tuning model technology, to automatically generate high-quality copy.

[0021] Furthermore, the automatic publishing module is used to publish the video to multiple platforms.

[0022] Furthermore, the moving wheel is a universal wheel equipped with a brake.

[0023] Furthermore, a limit block is installed on the right side of the cabinet door, a through slot is provided on the front side of the storage box, the cabinet door is installed in the through slot, and a limit slot is provided on the right side of the through slot, through holes arranged in concentric circles are provided inside the limit block and the limit slot, a movable shaft is inserted in the through hole, and heat dissipation holes are provided on the surrounding sides of the cabinet door and the storage box.

[0024] The beneficial effects of the present invention are as follows:

[0025] 1. The visual tracking module of the present invention includes a recognition unit, a transmission unit, a computing unit, a configuration unit, a data analysis unit and a learning unit, wherein the recognition unit supports data input from multiple sensors, such as cameras, lidars, etc., and the recognition unit includes target recognition, action recognition and posture estimation. Target recognition uses advanced deep learning algorithms, such as YOLO, FasterR-CNN, etc., to achieve high-precision target recognition; action recognition is used to recognize and track various human actions, such as walking, running, jumping, etc.; posture estimation analyzes the posture and motion trajectory of the human body through skeleton key point detection; the transmission unit is used to track the target in real time or near real time, and is suitable for application scenarios that require rapid response; the computing unit uses technologies such as GPU acceleration to improve processing speed and efficiency; the configuration unit is used to perform customized configuration according to different application scenarios to meet specific needs; the data analysis unit uses a large amount of historical data for training and optimization to improve the accuracy and robustness of the system; the learning unit supports online learning and model updating to continuously improve the performance and adaptability of the system.

[0026] 2. After the driving machine of the present invention is started, it can drive the driving wheel to perform circular motion. When the driving wheel rotates, it can drive the hydraulic cylinder, the support seat and the camera to adjust the angle. After the hydraulic cylinder is started, the height adjustment of the support seat and the camera can be achieved, thereby realizing automatic mirror movement, fixed-point shooting, multi-angle and multi-scene shooting, and thus being able to meet the use requirements under different conditions. It has strong practicality and good use effect.

[0027] 3. After the motor is started, the present invention can drive the movable column to move longitudinally along the side wall of the storage box under the action of the motor shaft, thereby achieving the purpose of moving the supporting tube, the cylinder and the display screen. Furthermore, after starting the cylinder, the height adjustment purpose of the display screen can be achieved, that is, with the cooperation of the motor, the movable column, the supporting tube and the cylinder, the height and longitudinal position of the display screen can be adjusted, and multi-angle adjustment is supported to ensure that the user can clearly and conveniently view the picture and data and operate the shooting task at any working angle.

[0028] 4. The automatic editing module of the present invention is used to achieve high-quality video editing through intelligent processing of the background system after the user completes shooting, such as sorting, editing, adding subtitles, special effects, color adjustment and audio processing. Regardless of whether the user has editing experience, he can easily complete professional-level editing; the copy generation module uses advanced AI large language model and self-designed prompt word system, combined with fine-tuning model technology, to automatically generate high-quality copy, support batch generation, greatly improve the efficiency and accuracy of copy creation, and under the action of the automatic publishing module, the edited video and copy can be sent to multiple platforms without tedious uploading operations, simplifying the publishing process and improving work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a three-dimensional schematic diagram of the present invention;

[0030] Figure 2 is a front cross-sectional view of the present invention;

[0031] Figure 3 is a three-dimensional schematic diagram of the automatic shooting module in the present invention;

[0032] Figure 4 It is a three-dimensional schematic diagram of the display mechanism in the present invention;

[0033] Figure 5 It is a working procedure diagram of the visual tracking module in the present invention;

[0034] Figure 6 It is a work flow chart of the present invention.

[0035] Figure numerals: 100, storage box; 200, automatic shooting module; 201, driving machine; 202, driving wheel; 203, transmission wheel; 204, hydraulic cylinder; 205, support base; 206, camera; 300, display mechanism; 301, motor; 302, moving column; 303, supporting tube; 304, cylinder; 305, display screen; 400, host; 500, switch; 600, bottom plate; 700, moving wheel; 800, cabinet door. DETAILED DESCRIPTION

[0036] To make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0037] See also Figure 1 - Figure 6 The present invention provides an intelligent shooting robot, comprising:

[0038] Storage box 100; the storage box 100 can support the automatic shooting module 200 and can also store various devices when not in use, thereby ensuring stability during movement or transportation. The storage box 100 is equipped with a high-power multi-hole PDU socket strip to meet the power requirements of the full set of workstation equipment. All lines are built-in, and users only need to connect an external power cord, which greatly simplifies the wiring operation.

[0039] The automatic shooting module 200 is installed at the center of the upper end of the storage box 100. The automatic shooting module 200 includes a driving machine 201. The circumferential surface of the upper end of the driving shaft of the driving machine 201 is sleeved with a driving wheel 202. The outer side of the driving wheel 202 is connected to a transmission wheel 203. The upper end of the transmission wheel 203 is installed with a hydraulic cylinder 204, a support seat 205 installed at the upper end of the hydraulic cylinder 204, and a camera 206 installed at the upper end of the support seat 205. The camera 206 has a built-in visual tracking module.

[0040] The storage box 100 can support the driving machine 201, the driving machine 201 can support the driving wheel 202, the driving wheel 202 can support the transmission wheel 203, the transmission wheel 203 can support the hydraulic cylinder 204, the hydraulic cylinder 204 can support the support seat 205, the support seat 205 can support the camera 206, and the camera 206 can shoot.

[0041] After starting the driving machine 201, it can drive the driving wheel 202 to perform circular motion. When the driving wheel 202 rotates, it can drive the hydraulic cylinder 204, the support seat 205 and the camera 206 to adjust the angle. After starting the hydraulic cylinder 204, the height adjustment purpose of the support seat 205 and the camera 206 can be achieved, thereby realizing automatic mirror movement, fixed-point shooting, multi-angle and multi-scene shooting, and thus being able to meet the use requirements under different conditions. It is highly practical and has a good use effect.

[0042] The display mechanism 300 is installed on the outer wall of the left side of the storage box 100. The display mechanism 300 includes a motor 301, a moving column 302 movably installed on the circumferential surface of the motor shaft of the motor 301, a supporting cylinder 303 installed on the outer side of the moving column 302, a cylinder 304 installed inside the supporting cylinder 303, and a display screen 305 installed on the upper end of the cylinder 304;

[0043] The cooperation of the motor 301 and its motor shaft can ensure the installation stability of the moving column 302, the action of the moving column 302 can ensure the installation stability of the supporting cylinder 303, the action of the supporting cylinder 303 can ensure the installation stability of the cylinder 304, and the action of the cylinder 304 can ensure the installation stability of the display screen 305.

[0044] After starting the motor 301, the movable column 302 can be driven to move longitudinally along the side wall of the storage box 100 under the action of the motor shaft, so as to achieve the purpose of moving the supporting tube 303, the cylinder 304 and the display screen 305. Furthermore, after starting the cylinder 304, the height adjustment purpose of the display screen 305 can be achieved, that is, with the cooperation of the motor 301, the movable column 302, the supporting tube 303 and the cylinder 304, the height and longitudinal position of the display screen 305 can be adjusted, supporting multi-angle adjustment, ensuring that the user can clearly and conveniently view the picture and data and operate the shooting task at any working angle.

[0045] The host 400 is installed inside the storage box 100, and the host 400 has built-in automatic editing module and text generation module; the automatic editing module can automatically edit the captured images and videos, and the text generation module can achieve the purpose of generating text.

[0046] The switch 500 is installed inside the storage box 100 and is located on the right side of the host 400. The switch 500 has a built-in automatic publishing module; under the function of the automatic publishing module, the edited video and text can be sent to multiple platforms.

[0047] The bottom plate 600 is installed below the storage box 100, and the moving wheels 700 are installed at the four corners of the lower end of the bottom plate 600;

[0048] The bottom plate 600 can support the storage box 100 and ensure the installation stability of the moving wheels 700. The bottom plate 600 and the components above it can be moved under the action of the moving wheels 700, thereby ensuring the use effect of the device.

[0049] The cabinet door 800 is detachably mounted on the front side of the storage box 100 via a movable shaft.

[0050] A magnetic block is installed on the left side of the cabinet door 800, and a sink is opened on the left side of the front of the storage box 100, and a magnetic connecting block is installed inside the sink. With the cooperation of the magnetic block and the magnetic connecting block, the installation stability of the left side of the cabinet door 800 can be guaranteed, and a handle is installed on the front of the left side of the cabinet door 800. The handle facilitates the opening and closing of the cabinet door 800, and the opening and closing of the cabinet door 800 facilitates the installation, maintenance and inspection of the equipment.

[0051] In this embodiment, preferably, a plurality of driving teeth are evenly installed on the circumferential surface of the driving wheel 202, and a plurality of transmission teeth are evenly opened inside the lower end of the transmission wheel 203, and the driving wheel 202 is meshed and connected with the transmission wheel 203; with the cooperation of the driving teeth and the transmission teeth, the transmission stability of the driving wheel 202 and the transmission wheel 203 can be guaranteed, thereby ensuring the rotation effect.

[0052] In this embodiment, preferably, the visual tracking module includes a recognition unit, a transmission unit, a computing unit, a configuration unit, a data analysis unit and a learning unit, wherein the recognition unit supports data input from multiple sensors, such as cameras, lidars, etc.; the transmission unit is used to track targets in real time or near real time, and is suitable for application scenarios that require fast response; the computing unit uses technologies such as GPU acceleration to improve processing speed and efficiency; the configuration unit is used to perform customized configuration according to different application scenarios to meet specific needs; the data analysis unit uses a large amount of historical data for training and optimization to improve the accuracy and robustness of the system; the learning unit supports online learning and model updates to continuously improve the performance and adaptability of the system.

[0053] In this embodiment, preferably, the recognition unit includes target recognition, action recognition and posture estimation. Target recognition utilizes advanced deep learning algorithms, such as YOLO, FasterR-CNN, etc., to achieve high-precision target recognition; action recognition is used to recognize and track various human actions, such as walking, running, jumping, etc.; posture estimation analyzes the posture and motion trajectory of the human body through skeleton key point detection.

[0054] In the present embodiment, preferably, a movable groove with arc-shaped structures at both ends is opened inside the left side of the storage box 100, and the right side of the movable column 302 is movably connected with the storage box 100 through the movable groove, and a threaded hole is opened inside the right side of the movable column 302, and the circumferential surface of the motor shaft of the motor 301 is provided with an external thread, and the movable column 302 is connected to the motor shaft through a thread; under the action of the movable groove, space can be provided for the movement of the movable column 302, and it also has the effect of limiting, and with the cooperation of the external thread and the threaded hole, the movable column 302 can be moved axially on the circumferential surface of the motor shaft.

[0055] In the present embodiment, preferably, a high-resolution display is selected for the display screen 305, and the display screen 305 is equipped with a matching mouse and keyboard. A positioning frame is installed on the outer side of the display screen 305, and the display screen 305 is installed in the positioning frame. The positioning frame can ensure the installation stability of the display screen 305, thereby ensuring the use effect of the display screen 305, and facilitating the user to monitor the shooting screen, make oral inscriptions, and adjust the shooting content in real time.

[0056] In this embodiment, the automatic editing module is preferably used to realize the high-quality video editing work after the user completes the shooting, through the intelligent processing of the background system, sorting, editing, adding subtitles, special effects, color adjustment and audio processing, etc., regardless of whether the user has editing experience, can easily complete professional-level editing; the copy generation module uses the advanced AI large language model and the independently designed prompt word system, combined with the fine-tuning model technology, to automatically generate high-quality copy, support batch generation, and greatly improve the efficiency and accuracy of copy creation.

[0057] In this embodiment, preferably, the automatic publishing module is used to publish videos to multiple platforms without the need for cumbersome uploading operations, thereby simplifying the publishing process and improving work efficiency.

[0058] In this embodiment, preferably, the moving wheel 700 is a universal wheel equipped with a brake, which can achieve the purpose of moving the device under the action of the universal wheel, and can achieve the purpose of positioning and fixing the entire device under the action of the brake, thereby ensuring the stability of the device during use.

[0059] In this embodiment, preferably, a limit block is installed on the right side of the cabinet door 800, a through slot is opened on the front side of the storage box 100, the cabinet door 800 is installed in the through slot, and a limit slot is opened on the right side of the through slot, through holes arranged in concentric circles are opened inside the limit block and the limit slot, the movable shaft is inserted in the through hole, and heat dissipation holes are opened on the surrounding sides of the cabinet door 800 and the storage box 100; the installation stability of the cabinet door 800 can be guaranteed by the action of the through slot, the movement stability of the limit block can be guaranteed by the cooperation of the limit slot and the movable shaft, and a good heat dissipation effect can be achieved by the action of the heat dissipation holes.

[0060] The working principle and use process of the present invention: when the device is used, the storage box 100 can support the automatic shooting module 200 and store the equipment when not in use, thereby ensuring stability during movement or transportation. The storage box 100 is equipped with a high-power multi-hole PDU socket to meet the power requirements of the full set of workstation equipment. All circuits are designed to be built-in, and the user only needs to connect an external power cord, which greatly simplifies the wiring operation. After starting the driving machine 201, the driving wheel 202 can be driven to perform circular motion. When the driving wheel 202 rotates, it can drive the hydraulic cylinder 204, the support seat 205 and the camera 206. The purpose of angle adjustment is achieved, and after the hydraulic cylinder 204 is started, the height adjustment purpose of the support base 205 and the camera 206 can be achieved, so that automatic mirror movement, fixed-point shooting, multi-angle and multi-scene shooting can be achieved, and then the use requirements under different conditions can be met. It has strong practicality and good use effect. The visual tracking module includes a recognition unit, a transmission unit, a calculation unit, a configuration unit, a data analysis unit and a learning unit. Among them, the recognition unit supports data input from multiple sensors, such as cameras, laser radars, etc. The recognition unit includes target recognition, action recognition and posture estimation. Target recognition uses advanced deep learning algorithms, such as YOLO, FasterR -CNN, etc., to achieve high-precision target recognition; action recognition is used to identify and track various human actions, such as walking, running, jumping, etc.; posture estimation analyzes the posture and motion trajectory of the human body through skeleton key point detection; the transmission unit is used to track targets in real time or near real time, which is suitable for application scenarios that require fast response; the computing unit uses technologies such as GPU acceleration to improve processing speed and efficiency; the configuration unit is used to customize configuration according to different application scenarios to meet specific needs; the data analysis unit uses a large amount of historical data for training and optimization to improve the accuracy and robustness of the system; the learning unit supports online learning and model updates , and continuously improve the performance and adaptability of the system; after starting the motor 301, the motor shaft can drive the moving column 302 to move longitudinally along the side wall of the storage box 100, so as to achieve the purpose of moving the supporting cylinder 303, the cylinder 304 and the display screen 305. Furthermore, after starting the cylinder 304, the height adjustment purpose of the display screen 305 can be achieved, that is, the cooperation of the motor 301, the moving column 302, the supporting cylinder 303 and the cylinder 304 can achieve the purpose of adjusting the height and longitudinal position of the display screen 305, and support multi-angle adjustment, ensuring that the user can clearly and conveniently view the picture and data and operate the shooting task at any working angle;The host 400 has built-in automatic editing module and copy generation module. The automatic editing module can automatically edit the images and videos that have been shot, and the copy generation module can achieve the purpose of generating copy. The automatic editing module is used to complete the high-quality video editing work through the intelligent processing of the background system after the user completes the shooting, and performs operations such as sorting, editing, adding subtitles, special effects, color adjustment and audio processing. Regardless of whether the user has editing experience or not, he can easily complete professional-level editing; the copy generation module uses the advanced AI large language model and the self-designed prompt word system, combined with the fine-tuning model technology, to automatically generate high-quality copy, support batch generation, and greatly improve the efficiency and accuracy of copy creation; the switch 500 has a built-in automatic publishing module, which can automatically publish the completed editing The videos and texts after the storage box 100 can be sent to multiple platforms without complicated uploading operations, which simplifies the publishing process and improves work efficiency. Under the action of the bottom plate 600, the storage box 100 can be supported and the installation stability of the moving wheel 700 can be ensured. Under the action of the moving wheel 700, the bottom plate 600 and the components above it can be moved, thereby ensuring the use effect of the device. A magnetic block is installed on the left side of the cabinet door 800, and a sink groove is opened on the left side of the front of the storage box 100, and a magnetic connection block is installed inside the sink groove. Under the cooperation of the magnetic block and the magnetic connection block, the installation stability of the left side of the cabinet door 800 can be ensured. A handle is installed on the front of the left side of the cabinet door 800, which is conducive to the opening and closing of the cabinet door 800. The opening and closing of the cabinet door 800 facilitates the installation, maintenance and inspection of the equipment. ;

[0061] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent photography robot, characterized in that: include: Storage box (100); An automatic shooting module (200) is installed at the center of the upper end of the storage box (100), and the automatic shooting module (200) comprises a driving machine (201), a driving wheel (202) is sleeved on the circumferential surface of the upper end of the driving shaft of the driving machine (201), a transmission wheel (203) is transmission-connected to the outer side of the driving wheel (202), a hydraulic cylinder (204) is installed at the upper end of the transmission wheel (203), a support seat (205) is installed at the upper end of the hydraulic cylinder (204), and a camera (206) is installed at the upper end of the support seat (205), and the camera (206) has a built-in visual tracking module; A display mechanism (300) is mounted on the outer wall of the left side of the storage box (100), and the display mechanism (300) comprises a motor (301), a moving column (302) movably mounted on the circumferential surface of the motor shaft of the motor (301), a supporting cylinder (303) mounted on the outer side of the moving column (302), a cylinder (304) mounted inside the supporting cylinder (303), and a display screen (305) mounted on the upper end of the cylinder (304); A host (400) is installed inside the storage box (100), and the host (400) has an automatic editing module and a text generation module built in; A switch (500) is installed inside the storage box (100) and is located on the right side of the host (400), and the switch (500) has an automatic publishing module built in; A bottom plate (600) installed below the storage box (100); Moving wheels (700) are installed at the four corners of the lower end of the bottom plate (600); The cabinet door (800) is detachably mounted on the front side of the storage box (100) via a movable shaft.

2. The intelligent photography robot according to claim 1, characterized in that: A plurality of driving teeth are evenly mounted on the circumferential surface of the driving wheel (202), a plurality of driving teeth are evenly arranged inside the lower end of the transmission wheel (203), and the driving wheel (202) is meshingly connected with the transmission wheel (203).

3. The intelligent photography robot according to claim 1, characterized in that: The visual tracking module includes a recognition unit, a transmission unit, a computing unit, a configuration unit, a data analysis unit and a learning unit, wherein the recognition unit supports data input from multiple sensors, such as cameras, lidars, etc.; the transmission unit is used to track targets in real time or near real time, and is suitable for application scenarios that require rapid response; the computing unit uses technologies such as GPU acceleration to improve processing speed and efficiency; the configuration unit is used to perform customized configuration according to different application scenarios to meet specific needs; the data analysis unit uses a large amount of historical data for training and optimization to improve the accuracy and robustness of the system; the learning unit supports online learning and model updates to continuously improve the performance and adaptability of the system.

4. The intelligent photography robot according to claim 3, characterized in that: The recognition unit includes target recognition, action recognition and posture estimation. The target recognition uses advanced deep learning algorithms, such as YOLO, FasterR-CNN, etc., to achieve high-precision target recognition; the action recognition is used to recognize and track various human actions, such as walking, running, jumping, etc.; the posture estimation analyzes the posture and motion trajectory of the human body through skeleton key point detection.

5. The intelligent photography robot according to claim 1, characterized in that: A movable groove with arc-shaped structures at both ends is provided inside the left side of the storage box (100), and the right side of the movable column (302) is movably connected with the storage box (100) through the movable groove, and a threaded hole is provided inside the right side of the movable column (302), and the circumferential surface of the motor shaft of the motor (301) is provided with an external thread, and the movable column (302) is connected to the motor shaft through a thread.

6. The intelligent photography robot according to claim 1, characterized in that: The display screen (305) is a high-resolution display, and the display screen (305) is equipped with a matching mouse and keyboard. A positioning frame is installed on the outside of the display screen (305), and the display screen (305) is installed in the positioning frame.

7. The intelligent photography robot according to claim 1, characterized in that: The automatic editing module is used to realize sorting, editing, adding subtitles, special effects, color adjustment and audio processing through intelligent processing of the background system after the user completes the shooting. The copy generation module uses the advanced AI large language model and the independently designed prompt word system, combined with the fine-tuning model technology, to automatically generate high-quality copy.

8. The intelligent photography robot according to claim 1, characterized in that: The automatic publishing module is used to publish videos to multiple platforms.

9. The intelligent photography robot according to claim 1, characterized in that: The moving wheel (700) is a universal wheel equipped with a brake.

10. The intelligent photography robot according to claim 1, characterized in that: A limit block is installed on the right side of the cabinet door (800), a through slot is provided on the front side of the storage box (100), the cabinet door (800) is installed in the through slot, and a limit slot is provided on the right side of the through slot, through holes arranged in a concentric circle are provided inside the limit block and the limit slot, a movable shaft is inserted into the through hole, and heat dissipation holes are provided on the peripheral sides of the cabinet door (800) and the storage box (100).

Images