A video conferencing apparatus
By using a video conferencing device with a pan-tilt camera and matrix microphone, combined with deep learning algorithms and wireless transmission, the problems of limited shooting range, unclear audio, complex connection and unstable signal of traditional video conferencing devices are solved, achieving efficient and convenient remote communication.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Utility models(China)
- Current Assignee / Owner
- SHENZHEN KERUN VISUAL TECH CO LTD
- Filing Date
- 2025-07-21
- Publication Date
- 2026-07-31
AI Technical Summary
Traditional video conferencing devices have shortcomings in image acquisition, audio processing, device connection and use, signal transmission and terminal adaptation, making it difficult to meet the needs of efficient, convenient and high-quality remote communication.
The video conferencing device uses a pan-tilt camera and matrix microphone, combined with deep learning algorithms to achieve full-area shooting and automatic speaker tracking. The array microphone automatically identifies the speaker's position and filters noise. The mainboard processes audio and video and transmits them wirelessly or via wired means.
It achieves full coverage of attendees, clearly captures speaker images and audio, simplifies device connections, improves communication efficiency and stability, and adapts to various terminal displays.
Smart Images

Figure CN224583222U_ABST
Abstract
Description
Technical Field
[0001] This utility model belongs to the technical field of conference devices, specifically relating to a video conferencing device. Background Technology
[0002] In today's era of increasingly prevalent digital office and remote collaboration, video conferencing, as an efficient communication tool, has been widely used in many fields such as corporate meetings, online education, and telemedicine. With the continuous evolution of technology, video conferencing devices have undergone a development process from basic to complex, and from single-function to diversified.
[0003] Traditional video conferencing devices are mostly built on dedicated hardware, with a multipoint control unit (MCU) at its core. The MCU is responsible for processing video and audio streams and coordinating communication between participants. Its working principle involves receiving video streams uploaded from each terminal, decoding them, combining them into multiple frames, re-encoding them, and transmitting them to other terminals or recording servers for recording and storage. This traditional architecture ensures a certain degree of stability and security in video quality. However, with the increasing diversity of application scenarios and the continuous improvement of user demands, the drawbacks of traditional video conferencing devices are becoming increasingly apparent.
[0004] Traditional cameras have several limitations in image acquisition. Firstly, their shooting range is often limited, making it difficult to fully cover all participants in a large conference room. For example, in a typical rectangular conference room layout with a large number of participants, those in corners or back rows may not be clearly captured, preventing remote participants from obtaining complete information about all attendees and impacting the comprehensiveness of communication. Secondly, traditional cameras struggle to automatically and accurately track and focus on a speaker. In seminars where multiple participants take turns speaking, if the camera cannot be switched to the speaker and zoomed in promptly, remote participants may find it difficult to clearly see the speaker's facial expressions, body language, and other important information, reducing the effectiveness of communication.
[0005] The audio capture capabilities are also inadequate. Ordinary microphones cannot automatically identify the source of sound based on the speaker's location in multi-person conference scenarios. In temporary meeting rooms converted from open-plan office environments, ambient noise (such as the operation of office equipment and the movement of people) easily gets mixed into the audio capture range, resulting in unclear audio. This not only affects the listening experience of local participants, but also makes it impossible for remote participants to hear the speaker's content clearly, severely interfering with the accurate transmission of meeting information.
[0006] From the perspective of device connectivity and ease of use, the installation and debugging process of traditional video conferencing equipment is cumbersome. It typically requires complex cabling, involving multiple cable connections such as video, audio, and network cables, placing high demands on the environment and layout of the installation site. In conference rooms converted from older buildings, cabling is difficult and costly, and may even be impossible to achieve an ideal cabling solution due to building structure limitations. Furthermore, the user interface of traditional equipment is complex, requiring professional personnel for operation and maintenance. For ordinary users, the learning curve is steep, reducing meeting preparation efficiency and limiting the application of video conferencing in a wider range of scenarios.
[0007] Furthermore, issues exist with audio and video signal transmission and terminal compatibility. Traditional video conferencing devices suffer from insufficient stability during signal transmission, making them susceptible to network fluctuations and signal interference. In remote areas with poor network conditions or small businesses with limited network bandwidth, video stuttering and audio interruptions frequently occur during meetings. Moreover, poor interface compatibility with display terminals (such as large screens) means that different brands and models of equipment may require different adapters or drivers, increasing the difficulty of device compatibility and leading to signal transmission interruptions or inability to display / play properly, thus affecting the smooth progress of the meeting.
[0008] Existing video conferencing devices have shortcomings in image acquisition, audio processing, device connection and use, signal transmission and terminal adaptation, making it difficult to meet the growing demand for efficient, convenient and high-quality remote communication. Utility Model Content
[0009] The purpose of this invention is to provide a video conferencing device that solves the following technical problems of existing video conferencing devices: limited shooting range, unable to fully cover all participants, and difficult to automatically track and focus on the speaker; ordinary microphones cannot automatically identify the speaker's position, are easily mixed with environmental noise, resulting in unclear audio acquisition; complex device connection wiring, cumbersome operation interface, inconvenient installation, debugging and use, requiring professional personnel to operate; poor stability of audio and video signal transmission, susceptible to interference, low compatibility with display terminals, and prominent compatibility issues.
[0010] To address the above technical problems, this utility model discloses a video conferencing device, comprising a cuboid housing, the housing including a front shell and a rear cover, wherein a speaker assembly, a video acquisition module, an audio acquisition module, an antenna, and a functional motherboard are mounted on the front shell, the video acquisition module including a lens housing, a lens rotating mount, and a camera assembly, the camera assembly being connected to the lens rotating mount via rotating shafts on the left and right sides, the lens rotating mount having a bearing seat, one end of the bearing seat being rotatably connected to the rotating shaft, and the other end of the bearing seat being connected to the video acquisition module mounting position on the front shell.
[0011] Preferably, the camera assembly includes a lens element, a camera module, a pressure plate, and a camera function board. The lens element covers the lens hole of the lens housing, the camera module is connected to the pressure plate, and the camera module extends into the acquisition hole of the lens housing and is fixedly connected.
[0012] Preferably, a rotating rubber ring is fitted on the rotating shaft, and the rotating rubber ring fills the gap between the rotating shaft and the bearing seat, so that the lens rotating seat rotates together with the lens housing.
[0013] Preferably, the lens rotating mount includes a left rotating mount and a right rotating mount, and the pressure plate is pressed against the left rotating mount and the right rotating mount for synchronous rotation of the left rotating mount and the right rotating mount at the same angle.
[0014] Preferably, the camera function board is fixedly installed on the top of the bracket, the bracket is inverted L-shape, the front of the bracket is fixedly connected to the front shell, and an HDMI adapter box is fixed between the front of the bracket and the front shell.
[0015] Preferably, the speaker assembly includes a horn and a grille that completely covers the front of the front housing.
[0016] Preferably, the rear cover is fixedly mounted with a phase inverter, a rocker switch, and an interface plate, and the interface plate is fixedly connected to the rear cover by staggered fixing posts and fixing plates.
[0017] Preferably, the speaker assembly further includes a cavity sealing foam that covers the corresponding position of the speaker on the inner side of the front housing.
[0018] Preferably, the audio acquisition module includes an array microphone and a microphone PCB board, and the audio acquisition module is fixedly installed on the outside of the front shell, facing the participants.
[0019] Preferably, the video acquisition module integrates a high-performance image processor and is equipped with a deep learning-based target detection and tracking algorithm. This module connects to a camera with a pan-tilt-zoom (PTZ) mechanism, enabling real-time analysis of the camera's captured images. It accurately identifies the positions of all participants in the conference room and adjusts the camera's tilt angle and PTP rotation via control commands to ensure the entire conference room is covered. When a speaker is detected through audio module linkage or lip movement recognition in the image, the camera automatically focuses and magnifies the speaker's area, solving the problems of limited shooting range and inability to automatically track and focus in traditional cameras. The audio acquisition module incorporates a matrix microphone control chip and an adaptive noise reduction circuit. After the multi-channel audio signals acquired by the array microphones are input, the speaker's position is quickly located using a sound source localization algorithm. Simultaneously, a noise reduction algorithm filters out environmental noise such as air conditioning sounds and keyboard noises, amplifying only the speaker's voice. The processed audio signal has significantly improved clarity, solving the problems of ordinary microphones being unable to identify sound sources and easily introducing noise.
[0020] Preferably, the hardware connections of this video conferencing device are as follows: the array microphone collects audio signals, the camera assembly collects video signals, and the array microphone and the camera assembly are connected to the microphone PCB board for digital signal processing; the camera assembly is connected to the camera function board for digital signal processing; the microphone PCB board is connected to the function motherboard via USB; the function motherboard uses a quad-core ARM Cortex-A53 2.0GHz processor, paired with an ARM Mali-450 GPU, a 4K@60fps video decoder, a 1080P@60fps video encoder, 2GB DDR RAM, and 16GB of memory, providing audio and video input; the function motherboard is connected to a speaker for audio output; the USB and HDMI interfaces are used to connect to a panel to display the transmitted video; and it is powered by DC 12V.
[0021] Compared with the prior art, the beneficial effects obtained by this utility model are: This utility model discloses a video conferencing device that uses a camera with pan-tilt camera control function, and can also manually adjust the camera angle by 15 degrees in both tilt and tilt. Combined with a matrix microphone with automatic recognition function, this video conferencing device is placed at the front of the conference room. The camera can capture all the participants in the conference room and automatically track the speaker and zoom in and focus as needed. The matrix microphone automatically recognizes and judges the different positions and directions of the speaker, automatically filters unwanted noise, and collects and amplifies the speaker's voice.
[0022] The video conferencing device disclosed in this utility model, after acquiring images and audio, is equipped with a wireless transmission module on its internal functional motherboard. The acquired signals are transmitted through a wireless antenna. At the other end, a decoding and receiving device with receiving function receives the wireless audio and video signals and connects them to a display terminal such as a large screen via a TP-C or HDMI data cable for display and playback.
[0023] When holding online meetings using this video conferencing device, both parties can clearly see and hear all the information from all participants. When a participant speaks, the focus can be on the speaker. This video conferencing device is easy to use and simple to connect, and can be used for online meetings or on-site teaching to improve communication efficiency. Attached Figure Description
[0024] Figure 1 This is a schematic diagram of the video conferencing device of this utility model.
[0025] Figure 2 This is an exploded view of the video conferencing device of this utility model.
[0026] Figure 3 This is a schematic diagram of the camera component of the video conferencing device of this utility model.
[0027] Figure 4 This is a hardware architecture block diagram of the video conferencing device of this utility model.
[0028] Reference numerals: 1-Housing; 2-Front housing; 3-Rear cover; 31-Phase inverter; 32-Rocker switch; 33-Interface board; 4-Speaker assembly; 41-Speaker; 42-Grate; 43-Sound chamber sealing foam; 5-Video acquisition module; 51-Lens housing; 511-Rotating shaft; 512-Rotating rubber ring; 52-Lens rotating mount; 521-Left rotating mount; 522-Right rotating mount; 523-Bearing seat; 53-Camera assembly; 531-Lens lens; 532-Camera module; 533-Pressure plate; 534-Camera function board; 535-Bracket; 6-Audio acquisition module; 61-Array microphone; 62-Microphone PCB board; 7-Antenna; 8-Functional mainboard. Detailed Implementation
[0029] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. However, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0030] Please see Figures 1 to 3This utility model discloses a video conferencing device, the structure of which is as follows: The device includes a cuboid housing 1, which is composed of a front shell 2 and a rear cover 3. A speaker assembly 4, a video acquisition module 5, an audio acquisition module 6, an antenna 7, and a functional motherboard 8 are integrated and mounted on the front shell 2.
[0031] The video acquisition module 5 is the core component for image acquisition and processing, comprising a lens housing 51, a lens rotating base 52, and a camera assembly 53. The camera assembly 53 is connected to the lens rotating base 52 via rotating shafts 511 on both sides. The lens rotating base 52 has a bearing seat 523, one end of which is rotatably connected to the rotating shaft, and the other end is connected to the video acquisition module mounting position on the front housing 2. Specifically, the camera assembly 53 consists of a lens element 531, a camera module 532, a pressure plate 533, and a camera function board 534. The lens element 531 covers the lens hole of the lens housing 51, protecting and optimizing light intake. The camera module 532 is connected to the pressure plate 533 and extends into the acquisition hole of the lens housing 51 for fixed connection. The camera function board 534 is fixedly mounted on top of an inverted L-shaped bracket 535, the front of which is fixedly connected to the front housing 2. An HDMI adapter box is also fixed between the front of the bracket 535 and the front housing 2.
[0032] To ensure the stability and smoothness of rotation, a rotating rubber ring 512 is fitted onto the rotating shaft 511. The rotating rubber ring 512 fills the gap between the rotating shaft 511 and the bearing seat 523, allowing the lens rotating base 52 and the lens housing 51 to rotate together. The lens rotating base 52 is divided into a left rotating base 521 and a right rotating base 522. The pressure plate 533 is pressed against the left rotating base 521 and the right rotating base 522 to ensure that the left rotating base 521 and the right rotating base 522 rotate synchronously at the same angle.
[0033] In addition, the video acquisition module 5 integrates a high-performance image processor and is equipped with a deep learning-based target detection and tracking algorithm. When connected to a camera with a pan-tilt mechanism, this module can analyze the camera's captured images in real time, accurately identify the positions of all participants in the conference room, and adjust the camera's tilt angle and pan-tilt rotation via control commands to ensure the shooting range covers the entire conference room. When a speaker is detected through audio module linkage or lip movement recognition in the image, the camera automatically focuses and magnifies the speaker's area, effectively solving the problems of limited shooting range and inability to automatically track and focus in traditional cameras. The pan-tilt mechanism has two built-in micro stepper motors, which control horizontal rotation and tilt adjustment respectively. The horizontal drive motor is connected to the bearing seat 523 of the lens rotating base 52 via a gear set. Each step of the motor rotation can drive the camera to rotate horizontally by 0.5°, achieving 360° rotation without blind spots. The tilt drive motor is linked to the rotation shaft 511 of the camera assembly 53 via a linkage structure, driving the camera to swing up and down around the rotation shaft. The adjustment range is strictly controlled within ±15° to avoid exceeding the protection range of the lens housing 51. The gimbal integrates a ribbon cable interface at its bottom, connecting to the mainboard 8 via the camera function board 534. This interface receives angle control commands from the mainboard using PWM pulse signals and also provides feedback on the current angle and position using a Hall sensor. Simultaneously, the ribbon cable has a built-in power supply line to provide a stable voltage to the stepper motor, ensuring fast drive response. Through this structural design, the gimbal can achieve automated tracking and full-area coverage via electric drive, while also being compatible with manual adjustment needs. It provides flexible and reliable mechanical support for the video acquisition module 5 and is a key component for achieving "no blind spots in shooting and high-precision tracking."
[0034] The audio acquisition module 6 includes an array microphone 61 and a microphone PCB board 62, which are fixedly mounted on the outside of the front shell 2 and face the participants for efficient sound acquisition. The audio acquisition module 6 incorporates a matrix microphone control chip and an adaptive noise reduction circuit. When the multi-channel audio signal acquired by the array microphone 61 is received, the speaker's location can be quickly locked using a sound source localization algorithm. Simultaneously, a noise reduction algorithm is activated to filter environmental noise such as air conditioning noise and keyboard noise, amplifying only the speaker's voice. The processed audio signal has significantly improved clarity, solving the problem of ordinary microphones being unable to identify sound sources and easily introducing noise.
[0035] The loudspeaker assembly 4 consists of a speaker 41, a grille 42, and a sound cavity sealing foam 43. The grille 42 completely covers the front of the front shell 2, providing protection without affecting sound propagation; the sound cavity sealing foam 43 covers the inner side of the front shell 2 at the position corresponding to the speaker 41, which can enhance the sealing of the sound cavity and improve the sound quality.
[0036] A phase inverter 31, a rocker switch 32, and an interface plate 33 are fixedly installed on the rear cover 3. The interface plate 33 is fixedly connected to the rear cover 3 by staggered fixing posts and fixing plates, ensuring the stability of the connection.
[0037] The hardware connections of this video conferencing device are as follows: the array microphone 61 collects audio signals, the camera assembly 53 collects video signals, and the array microphone 61 is connected to the microphone PCB board 62 for digital signal processing; the camera assembly 53 is connected to the camera function board 534 for digital signal processing; the microphone PCB board and the camera function board 534 are connected to the function motherboard 8 via USB. The function motherboard 8 uses a quad-core ARM Cortex-A53 2.0GHz processor, paired with an ARM Mali-450 GPU, a 4K@60fps video decoder, a 1080P@60fps video encoder, 2GB DDR RAM, and 16GB of memory. The video acquisition module 5 and the audio acquisition module 6 provide audio and video inputs. The function motherboard 8 is connected to the speaker assembly 4 for audio output, uses USB and HDMI interfaces for connecting to a panel to display the transmitted video, and is powered by DC 12V.
[0038] Through the coordinated operation of the above structures, this video conferencing device can achieve high-quality audio and video acquisition, processing, and transmission, meeting the needs of online meetings or on-site teaching scenarios and effectively improving communication efficiency.
[0039] During use, the video conferencing device operates around audio and video acquisition, processing, transmission, and interaction. After turning on the rocker switch 32 on the back cover 3 and powering on the device, the mainboard 8 starts up and completes initialization. At this time, the camera function board 534 of the video acquisition module 5 and the microphone PCB board 62 of the audio acquisition module 6 enter working mode. The lens housing 51 and the lens swivel mount 52 reset to their initial horizontal centering angle and 0° tilt angle. The antenna 7 initiates wireless signal search, and the HDMI adapter box completes the signal path establishment with the mainboard.
[0040] Full-area screen coverage: The high-performance image processor of the video acquisition module 5 is activated, and the camera module 532 acquires the initial image through the lens 531. The processor analyzes the distribution of attendees in the image based on a deep learning algorithm and sends control commands to the gimbal: if individuals at the edge are detected not within the field of view, the bearing seat 523 of the lens rotating base 52 drives the rotating shaft 511 to rotate, achieving horizontal adjustment; simultaneously, through the synchronous linkage of the left rotating base 521 and the right rotating base 522, the pressure plate 533 ensures rotation at the same angle, adjusting the camera's tilt angle within ±15° until all attendees are covered. The rotating rubber ring 512 reduces rotational friction during this process, ensuring smooth angle adjustment.
[0041] Speaker Tracking and Focusing: When the audio acquisition module 6 captures the speaker's voice through the array microphone 61, the sound source localization algorithm locks the speaker's location and sends the coordinate information to the video acquisition module 5. The image processor, combined with lip movement recognition technology, reconfirms the speaker's position, drives the pan-tilt unit to rotate rapidly, and aligns the camera module 532 with the speaker; simultaneously, it controls the lens focal length to magnify and focus on the speaker's area, ensuring that the remote terminal clearly displays the speech details. If the speaker moves, the processor updates the position information in real time, achieving continuous tracking through dynamic adjustment of the rotating shaft and bearing seat.
[0042] Sound source acquisition and noise reduction: The array microphone 61 synchronously acquires multi-channel audio signals and transmits them to the matrix microphone control chip on the microphone PCB board 62. The chip locates the speaker's position through phase difference analysis and activates the adaptive noise reduction circuit: it filters environmental noise such as air conditioner noise and keyboard noise, retaining only the speaker's voice signal; at the same time, it enhances the sound intensity through a gain amplification circuit to avoid volume attenuation when speaking at a distance.
[0043] Audio output: The processed audio signal is transmitted in two paths: one path is transmitted through the functional motherboard 8 to the speaker 41 of the speaker assembly 4, and played through the mesh cover 42 for local participants to listen to. The sound cavity sealing foam 43 reduces sound leakage and improves sound quality; the other path enters the wireless transmission module synchronously with the video signal.
[0044] Signal Encoding and Transmission: The mainboard 8 encodes the processed audio and video signals. Generally, the video uses H.265 format and the audio uses AAC format, which is then wirelessly transmitted to the decoding and receiving device via antenna 7. If a wired connection is required, it can be directly connected to the display terminal via the TYPE-C or HDMI interface on the interface board 33 on the rear cover, through an HDMI adapter box.
[0045] Remote signal feedback: The audio and video signals of the remote terminal are transmitted back to the main board 8 via wireless or wired means. The audio signal drives the speaker 41 to play, and the video signal is transmitted to the display terminal through the camera function board 534 to realize two-way real-time communication. During this process, the bass reflex tube 31 of the back cover 3 optimizes the low-frequency response of the speaker and improves the reproduction of the remote sound.
[0046] After the meeting, turn off the rocker switch 32. Before the device loses power, the camera will automatically reset to its initial angle, and all modules will enter standby mode. If a temporary interruption is needed, the external connection can be disconnected via the interface board 33, or the transmission can be paused directly through the software interface of the main board.
[0047] Through the above process, the device achieves full-link automation from audio and video acquisition and intelligent processing to two-way transmission, ensuring clear images, pure sound, and smooth interaction during the meeting, meeting the requirements of efficiency and stability for remote communication.
[0048] The above examples are merely specific embodiments of this utility model. Obviously, this utility model is not limited to the above embodiments, and many similar modifications are possible. All variations that can be directly derived or conceived by those skilled in the art from the content disclosed in this utility model should be considered within the scope of protection of this utility model.
Claims
1. A video conferencing apparatus, characterized by comprising: The housing includes a rectangular shell (1), which includes a front shell (2) and a rear cover (3). The front shell (2) is equipped with a speaker assembly (4), a video acquisition module (5), an audio acquisition module (6), an antenna (7), and a functional motherboard (8). The video acquisition module (5) includes a lens housing (51), a lens rotator (52), and a camera assembly (53). The camera assembly (53) is connected to the lens rotator (52) via rotating shafts (511) on the left and right sides. The lens rotator (52) has a bearing seat (523). One end of the bearing seat (523) is rotatably connected to the rotating shaft, and the other end of the bearing seat (523) is connected to the video acquisition module mounting position on the front shell (2).
2. The video conferencing device according to claim 1, characterized in that, The camera assembly (53) includes a lens (531), a camera module (532), a pressure plate (533), and a camera function board (534). The lens (531) covers the lens hole of the lens housing (51). The camera module (532) is connected to the pressure plate (533). The camera module (532) extends into the acquisition hole of the lens housing (51) and is fixedly connected.
3. The video conferencing device according to claim 1, characterized in that, A rotating rubber ring (512) is fitted on the rotating shaft (511). The rotating rubber ring (512) fills the gap between the rotating shaft (511) and the bearing seat (523), so that the lens rotating seat (52) rotates together with the lens housing (51).
4. The video conferencing device according to claim 2, characterized in that, The lens rotating mount (52) includes a left rotating mount (521) and a right rotating mount (522). The pressure plate (533) is close to the left rotating mount (521) and the right rotating mount (522) for synchronous rotation of the left rotating mount (521) and the right rotating mount (522) at the same angle.
5. The video conferencing device according to claim 2, characterized in that, The camera function board (534) is fixedly installed on the top of the bracket (535). The bracket (535) is inverted L-shaped. The front of the bracket (535) is fixedly connected to the front shell (2), and an HDMI adapter box is fixed between the front of the bracket (535) and the front shell (2).
6. The video conferencing device according to claim 1, characterized in that, The speaker assembly (4) includes a horn (41) and a grille (42) that completely covers the front of the front shell (2).
7. The video conferencing device according to claim 4, characterized in that, The rear cover (3) is fixedly installed with a phase inverter (31), a rocker switch (32) and an interface plate (33). The interface plate (33) is fixedly connected to the rear cover (3) through staggered fixing posts and fixing plates.
8. The video conferencing device according to claim 6, characterized in that, The speaker assembly (4) also includes a cavity sealing foam (43) that covers the corresponding position of the speaker (41) inside the front housing (2).
9. The video conferencing device according to claim 2, characterized in that, The audio acquisition module (6) includes an array microphone (61) and a microphone PCB board (62). The audio acquisition module (6) is fixedly installed on the outside of the front shell (2) and faces the participants. The video acquisition module (5) integrates a high-performance image processor and is equipped with a target detection and tracking algorithm based on deep learning. The video acquisition module (5) is connected to a camera assembly (53) with a pan-tilt unit. It analyzes the images captured by the camera in real time, accurately identifies the positions of all participants in the conference room, and adjusts the camera tilt angle and pan-tilt unit rotation through control commands to ensure that the shooting range covers the entire conference room. The audio acquisition module (6) has a built-in matrix microphone control chip and an adaptive noise reduction circuit. After the multi-channel audio signals collected by the array microphone (61) are transmitted, the speaker's position is quickly locked by the sound source localization algorithm. At the same time, the noise reduction algorithm is activated to filter environmental noise and only the speaker's voice is amplified.
10. The video conferencing device according to claim 9, characterized in that, The hardware connections of the video conferencing device are as follows: the array microphone (61) collects audio signals, the camera assembly (53) collects video signals, the array microphone (61) is connected to the microphone PCB board (62) for digital signal processing; the camera assembly (53) is connected to the camera function board (534) for digital signal processing; the microphone PCB board (62) and the camera function board (534) are connected to the function motherboard (8) via USB. The function motherboard (8) uses a quad-core ARM Cortex-A53 2.0GHz processor, equipped with an ARM Mali-450 GPU, a 4K@60fps video decoder, a 1080P@60fps video encoder, 2GB DDR RAM, and 16GB DRAM. The function motherboard (8) is connected to the speaker assembly (4) for audio output, uses a USB interface and an HDMI interface for connecting to the panel to display the transmitted video, and is powered by DC 12V.