阅读说明：本技术 一种嵌入式数字会议方法以及会议系统 (Embedded digital conference method and conference system ) 是由 王芝彬 徐克兴 张志军 于 2019-09-25 设计创作，主要内容包括：本发明公开了一种嵌入式数字会议方法以及会议系统,所述会议方法中终端子板话筒的信号发送到主板进行合成,且合成后的信号分别分发到每个终端的扬声器中；发言者扬声器无自己的发言信号,当某终端处于发言状态时,则所述主板将合成的信号减去发言的终端的话筒信号后发送到发言的终端的扬声器中。所述会议系统包括若干个终端,所述解码模块、校准电路的输出端分别与D/A变换模块、拟合模块连接,所述A/D变换模块、拟合模块的输出端分别与抵消模块连接,且A/D变换模块的输出端还连接校准电路。本发明可以实现自适应回声对消、空间自适应,在发言时,所对应区域的扬声器音量自动减小,有效减少正反馈形成的啸叫,实现高保真音质。(The invention discloses an embedded digital conference method and a conference system, wherein in the conference method, signals of a terminal sub-board microphone are sent to a main board for synthesis, and the synthesized signals are respectively distributed to a loudspeaker of each terminal; the speaker of the speaker does not have a speaking signal of the speaker, and when a certain terminal is in a speaking state, the main board subtracts the microphone signal of the speaking terminal from the synthesized signal and sends the signal to the speaker of the speaking terminal. The conference system comprises a plurality of terminals, the output ends of the decoding module and the calibration circuit are respectively connected with the D/A conversion module and the fitting module, the output ends of the A/D conversion module and the fitting module are respectively connected with the offset module, and the output end of the A/D conversion module is also connected with the calibration circuit. The invention can realize self-adaptive echo cancellation and space self-adaptation, and when speaking, the volume of the loudspeaker in the corresponding area is automatically reduced, thereby effectively reducing howling formed by positive feedback and realizing high fidelity tone quality.)

1. An embedded digital conference method is characterized in that signals of a terminal sub-board microphone are sent to a main board for synthesis, and the synthesized signals are respectively distributed to a loudspeaker of each terminal; the speaker of the speaker does not have a speaking signal of the speaker, and when a certain terminal is in a speaking state, the main board subtracts the microphone signal of the speaking terminal from the synthesized signal and sends the signal to the speaker of the speaking terminal.

2. The method of claim 1, wherein when a terminal is in a speaking state, the speakers of the terminal and its neighboring terminals do not sound the speaking signal.

3. The embedded digital conference method according to claim 1 or 2, wherein the terminal filters out the sound signal of the speaker received by the microphone and retains the sound signal outside the speaker; carrying out convolution transformation on a signal input by a loudspeaker power amplifier and a calibration signal so as to obtain a sound signal of the loudspeaker received by a microphone; the calibration signal is obtained by a calibration circuit.

4. The embedded digital conference method according to claim 3, wherein, during the circuit calibration, the calibration circuit sends a long-period narrow pulse to the speaker, the period of the narrow pulse is 100mS, and the pulse width is the sampling period of the sound signal; the microphone receives narrow pulse sound emitted by the loudspeaker in the whole period, and the calibration signal with low noise is obtained after long-time accumulative averaging.

5. An embedded digital conference system is characterized by comprising a plurality of terminals, wherein each terminal comprises a decoding module, a D/A conversion module, an audio amplification module, a loudspeaker, a fitting module, a calibration circuit, a cancellation module, an A/D conversion module and a microphone; the output ends of the decoding module and the calibration circuit are respectively connected with the D/A conversion module and the fitting module, and the output end of the D/A conversion module is connected with the loudspeaker through the audio amplification module; the output end of the microphone is connected with the A/D conversion module through the audio amplification module, the output ends of the A/D conversion module and the fitting module are respectively connected with the offset module, and the output end of the A/D conversion module is further connected with the calibration circuit.

6. The embedded digital conference system of claim 5, further comprising a central controller, wherein said terminal further comprises a terminal processor, and wherein said central controller is connected to said terminal processor; the central controller is used for receiving the uplink signals of the terminals, synthesizing and packaging the uplink signals, local display data and broadcast audio data, and distributing the uplink signals to the terminal processors; and the terminal processor is used for uploading the local microphone audio data, receiving the downlink data distributed by the central controller, processing the downlink data and then sending the processed downlink data to the loudspeaker and the display.

7. The embedded digital conference system of claim 6, wherein said terminals are arranged in a rectangular array, adjacent terminals in each row are connected to each other, and the terminal processor of the first terminal in each row is connected to the central controller.

Technical Field

The invention belongs to the technical field of audio data processing, and particularly relates to an embedded digital conference method and a conference system.

Background

The existing local conference system has various forms, the local conference mainly adopts an analog system, a broadcasting form, namely a speaking seat can speak, the number of the speaking seats is limited, a conference hall can be amplified and shared through a large sound box, the system is difficult to realize interaction, multi-person discussion cannot be well realized, and noise accumulation is the greatest inherent defect.

In an indoor sound amplifying system, sound signals played by a loudspeaker after sound amplification are reflected by a room, picked by a microphone and amplified and then played for multiple cycles, and in the closed loop system, when a certain condition is met, an echo howling phenomenon can be generated. For example: when A and B carry out space-isolated communication through a loudspeaker and a microphone, A transmits sound to the loudspeaker of B through the microphone to produce sound, the produced sound is transmitted to the loudspeaker of A through the microphone of B to produce sound, the sound is transmitted to the loudspeaker of B through the microphone of A to produce sound, and the circulation is repeated, so that howling is caused. The howling phenomenon affects the progress of the conference and brings obstacles to the communication with the participants.

On the other hand, a software video conference system realized by combining software based on a server and a high-performance PC has the defects that the processing capacity of a CPU (central processing unit) of the computer is limited, the operation requirement of audio and video coding and decoding cannot be met, partial definition is sacrificed or too high bandwidth is occupied, and the real-time performance is poor due to time delay caused by soft coding and decoding.

The PC-based architecture is easy to be attacked by viruses, and a large number of plug-in cards and peripheral matching equipment are adopted, so that the integration level is low, the number of fault points is large, and the fault troubleshooting is difficult. Therefore, the safety and stability of the system cannot be guaranteed. Less suitable for certain conference sites, such as: where security requirements are high.

Disclosure of Invention

The invention aims to provide an embedded digital conference method, aiming at eliminating howling in a conference system. Another objective of the present invention is to provide an embedded digital conference system, which is intended to implement adaptive echo cancellation and has better practicability.

The invention is mainly realized by the following technical scheme: a method for embedded digital conference, the signal of the terminal sub-board microphone is sent to the main board to be synthesized, and the synthesized signal is distributed to the loudspeaker of each terminal; the speaker of the speaker does not have a speaking signal of the speaker, and when a certain terminal is in a speaking state, the main board subtracts the microphone signal of the speaking terminal from the synthesized signal and sends the signal to the speaker of the speaking terminal.

In order to better implement the present invention, further, when a terminal is in a speech state, the speakers of the terminal itself and its neighboring terminals do not sound the speech signal.

In order to better implement the invention, further, the terminal filters the sound signal of the speaker received by the microphone and retains the sound signal outside the speaker; carrying out convolution transformation on a signal input by a loudspeaker power amplifier and a calibration signal so as to obtain a sound signal of the loudspeaker received by a microphone; the calibration signal is obtained by a calibration circuit. Namely an echo cancellation algorithm.

In order to better implement the invention, further, during circuit calibration, the calibration circuit sends a long-period narrow pulse to the speaker, the period of the narrow pulse is 100mS, and the pulse width is a sound signal sampling period; the microphone receives narrow pulse sound emitted by the loudspeaker in the whole period, and the calibration signal with low noise is obtained after long-time accumulative averaging.

The invention is mainly realized by the following technical scheme: an embedded digital conference system comprises a plurality of terminals, wherein each terminal comprises a decoding module, a D/A conversion module, an audio amplification module, a loudspeaker, a fitting module, a calibration circuit, a cancellation module, an A/D conversion module and a microphone; the output ends of the decoding module and the calibration circuit are respectively connected with the D/A conversion module and the fitting module, and the output end of the D/A conversion module is connected with the loudspeaker through the audio amplification module; the output end of the microphone is connected with the A/D conversion module through the audio amplification module, the output ends of the A/D conversion module and the fitting module are respectively connected with the offset module, and the output end of the A/D conversion module is further connected with the calibration circuit.

In order to better implement the present invention, the terminal further comprises a central controller, the terminal further comprises a terminal processor, and the central controller is connected with the terminal processor; the central controller is used for receiving the uplink signals of the terminals, synthesizing and packaging the uplink signals, local display data and broadcast audio data, and distributing the uplink signals to the terminal processors; and the terminal processor is used for uploading the local microphone audio data, receiving the downlink data distributed by the central controller, processing the downlink data and then sending the processed downlink data to the loudspeaker and the display.

In order to better implement the present invention, further, the terminals are arranged in a rectangular array, adjacent terminals in each row are connected to each other, and the terminal processor of the first terminal in each row is connected to the central controller respectively.

The invention has the beneficial effects that:

(1) the invention can realize self-adaptive echo cancellation, space self-adaptation (when a terminal is in a speaking state, the speaker of the terminal and the adjacent terminal do not speak the speaking signal), and multi-person discussion.

(2) When speaking, the volume of the loudspeaker in the corresponding area is automatically reduced, thereby effectively reducing howling formed by positive feedback and realizing high fidelity tone quality. The invention adopts low-noise large-dynamic audio Codec and high-speed FPGA to complete all algorithms, has higher integration level, adopts independent network protocol, and ensures good security of the system by radiation interception prevention of wired system baseband transmission.

(3) The conference system is based on an embedded architecture, audio and video processing, network communication and various conference functions are realized through the FPGA and embedded software and by means of professional hardware equipment terminals and the like, and the embedded architecture has good stability and safety.

(4) The invention mainly aims at places with higher confidentiality requirements, such as national defense, government and other conference places, is completely different from the traditional conference system in technology, fully considers the practical situation of conference application, embodies innovative design concept, advanced signal processing capability and high confidentiality and security from system architecture to specific realization, and really realizes multi-person interactive discussion and CD tone quality.

(5) The invention uses independent network protocol and algorithm, and personalized design enables each participant to have an independent terminal, thereby being convenient for use; the conference participants are confronted with each other by using the low-power loudspeaker and the hidden microphone in a scientific and reasonable layout without shielding, the conference participants are given sufficient freedom in space, and the comfort level of speaking, viewing and listening is improved.

(6) The invention adopts FPGA chip integration, comprising: processing capabilities such as a sound console, a pressure limiter, a self-adaptive echo suppressor, a space self-adaptive algorithm and the like; when speaking, the volume of the loudspeaker in the corresponding area is automatically reduced, thereby effectively reducing howling formed by positive feedback and realizing high fidelity tone quality. In addition, the system can be expanded to realize the functions of electronic table cards, electronic white boards, picture scrolling, streaming media playing and the like.

Drawings

FIG. 1 is a functional block diagram of a motherboard;

fig. 2 is a functional block diagram of a daughter board.

Detailed Description