CN100505041C - Sound signal collecting and processing system and method thereof - Google Patents

Sound signal collecting and processing system and method thereof Download PDF

Info

Publication number
CN100505041C
CN100505041C CN 200610151451 CN200610151451A CN100505041C CN 100505041 C CN100505041 C CN 100505041C CN 200610151451 CN200610151451 CN 200610151451 CN 200610151451 A CN200610151451 A CN 200610151451A CN 100505041 C CN100505041 C CN 100505041C
Authority
CN
China
Prior art keywords
sound
module
sensing element
sound signal
acoustic sensing
Prior art date
Application number
CN 200610151451
Other languages
Chinese (zh)
Other versions
CN101140760A (en
Inventor
胡俭波
Original Assignee
联想移动通信科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 联想移动通信科技有限公司 filed Critical 联想移动通信科技有限公司
Priority to CN 200610151451 priority Critical patent/CN100505041C/en
Publication of CN101140760A publication Critical patent/CN101140760A/en
Application granted granted Critical
Publication of CN100505041C publication Critical patent/CN100505041C/en

Links

Abstract

本发明公开了一种声音信号采集处理系统及方法,其中,该系统包括声音信号采集模块、声音信号预处理模块、波束形成模块、D/A转换模块和噪音消除模块,用于获取波束形成模块输出的波束中频率与第一频谱分量频率相同的部分,并发送给D/A转换模块,所述第一频谱分量在声音传感元件拾音面的声压差处于声压差预定值范围;所述声音信号采集模块包括多个声音传感元件,不同声音传感元件的拾音面与需要采集的声源的距离不同。 The present invention discloses a collected sound signal processing system and method, wherein the system includes a sound signal acquisition module, a sound signal pre-processing module, a beam forming module, D / A conversion module and a noise canceling means for obtaining a beam forming module frequency of the output beam in the same portion of the first component of the frequency spectrum, and sent to D / a converter module, the first spectral components in the sound pressure of the acoustic sensing element is in the pickup surface sound pressure difference predetermined value range; the sound signal acquisition module comprises a plurality of acoustic sensing elements, different distances from the sound source of the pickup surface to be collected different acoustic sensing element. 本发明通过将声音传感元件同轴同向,且间隔一定距离设置,这样在波束形成后,将波束内声压差不处于预定值范围的频谱分量进行滤除,有效地消除了波束内和波束外的噪音的影响,提高了信号处理的质量。 By the present invention the acoustic sensing element coaxially in the same direction, and arranged at a distance, so that after beam forming, the pressure within the acoustic beam is not in the predetermined range of values ​​of the spectral components were filtered off, effectively eliminating the beam and Effect of noise outside the beam, improve the quality of signal processing.

Description

一种声音信号采集处理系统及方法 A sound signal acquisition and processing system and method

技术领域 FIELD

本发明涉及声音信号的处理,特别是可消除背景噪音的声音信号采集装置、声音信号采集处理装置及方法。 The present invention relates to audio signal processing, especially the background noise sound signal acquisition device, the collected sound signal processing apparatus and method can be eliminated. 背景技术 Background technique

目前手机、PDA、电话机、带有声音信号采集装置的耳机等都需要用到声音信号采集装置,利用该声音信号采集装置采集需要发送的语音信号后发送给对方。 At present mobile phone, PDA, telephone with headset sound signal acquisition device and so need to use the sound signal acquisition device, using voice signals of the sound signal acquisition device you want to send to send to each other later.

然而,当使用者处于一个嘈杂环境中,背景噪声分贝很高,声音信号采集装置采集到的信号中就会包括背景噪音,将该采集到的信号解码输出时会很难听清原来的话音。 However, when the user is in a noisy environment, the background noise is high decibel, the audio signal acquisition device will include a signal to background noise, when the output of the decoded signal will be collected hard to hear the original voice.

从数据分析可以获知,只有使用者输出的语音高于背景噪声15dB以上, 才能从还原后的声音中听清使用者的语音。 Can be learned from the data analysis, only the user's voice above the background noise output of 15dB or more, the user can hear the voice from the sound of the restored. 使用者的语音低于背景噪声15dB 以上时,如果想要还原的信号中能清晰的辨别出使用者的语音,通常情况下, 通话者只能高声大喊,其实这样效果很有限, 一是由于背景噪声分贝已经很高的原因,使用者很难超过其15dB,同时,声音信号采集装置的增益有限, 使用者声音过大将会造成失真,还原出的信号中仍然可能无法清晰的分辨出使用者的话音。 When the user's voice is lower than 15dB above background noise, if you want to restore the signal can clearly identify the user's voice, usually, the call can only shout loudly, in fact, the effect is very limited, first, due to the background noise decibel is already high, the user is difficult to exceed its 15dB, while the gain of the sound signal acquisition device is limited, the user voice is too large will cause distortion, a reduction in the signal may still not clearly distinguish the use voice persons.

目前,声音信号采集装置包括一种双麦克风方案,利用波束形成算法消除波束外的噪声,其麦克风采用如图l所示的并排方式。 At present, the sound signal collection means comprises a dual microphone embodiment, the use of beamforming algorithms remove noise outside the beam, which uses a microphone side by side as shown in Figure l.

图1所示的并排方式的麦克风阵列在实现波束形成的算法采用基于频域LMS (Least Mean Square,最小均方)的自适应波束形成算法,其处理过程如下所示,包括: The microphone array of side by side in FIG. 1 algorithm beamforming algorithm based on the frequency domain is formed using LMS (Least Mean Square, Least Mean Square) adaptive beam, which follows the process, comprising:

步骤Sl,将麦克风阵列采集到的声音信号转换成数字语音信号4")后, 对输入信号:c(w)进行FFT (Fast Fourier Transform,快速傅立叶变换),用矩阵表示为K")-約(n),其中:W为频域变换矩阵,酉矩阵,为:对阵列接收信号进行N点FFT中,N为阵列中麦克风数,如果麦克风数是不为2的整数次幂,则采用补零的办法。 Step Sl is, the microphone array collected converts the sound signal into a digital voice signal 4 "), then the input signal: c (w) for FFT (Fast Fourier Transform, Fast Fourier Transform), expressed as K matrix") - about (n-), wherein: W is a frequency domain transform matrix, the unitary matrix, as follows: array reception signal N-point FFT, N is the array number of microphones, if the number of microphones is not an integer power of 2, is used up approach zero.

步骤S2,利用LMS算法对经过FFT的输入信号进行处理: Step S2, the LMS algorithm using the input signal through the FFT process is performed:

_K") = rr (")/•(") 其中:V为LMS算法中的权向量。 _K ") = rr (") / • ( ") where: V is the weight vector LMS algorithm.

其中:d为训练序列。 Where: d is the training sequence.

F(" +1) = F (") + 2ae(")r * (p) 其中:a为学习步长。 F ( "+1) = F (") + 2ae ( ") r * (p) where: a is the learning step.

增加样本,循环执行上述的步骤,最终得到波束,该最终得到的波束消除了波束外的噪音。 Increasing the sample, the above step cycle, finally obtained beam, the resulting beam to eliminate noise outside the beam.

然而上述的基于频域LMS (Least Mean Square,最小均方)的自适应波束形成算法无法消除在波束内的噪声。 However, the above frequency domain is formed based on the LMS (Least Mean Square, Least Mean Square) adaptive beam algorithm can not remove noise in the beam. 发明内容 SUMMARY

本发明的目的在于提供一种声音信号采集处理系统及方法,同时消除消除波束内和波束外的噪音。 Object of the present invention is to provide a sound signal acquisition and processing systems and methods, while eliminating external noise cancellation within the beam and beam.

为了实现上述目的,本发明提供了一种声音信号采集处理系统,包括声音信号采集模块、声音信号预处理模块、波束形成模块、D/A转换模块和噪音消除模块,该噪音消除模块用于获取波束形成模块输出的波束中频率与第一频谱分量频率相同的部分,并发送给D/A转换模块,所述第一频谱分量在声音传感元件拾音面的声压差处于声压差预定值范围; To achieve the above object, the present invention provides a sound signal acquisition and processing system, comprising a sound signal acquisition module, a sound signal pre-processing module, a beam forming module, D / A conversion module and a noise cancellation module, configured to obtain the noise cancellation module beamforming module output beam frequency component of the same portion of the first frequency spectrum, and sends the D / a converter module, the first spectral components in the sound pressure of the acoustic sensing element pickup surface is a predetermined sound pressure difference value range;

所述声音信号采集模块包括多个声音传感元件,不同的声音传感元件的拾音面与需要采集的声源的距离不同。 The sound from the signal acquisition module comprises a plurality of acoustic sensing elements, different sound pickup surface of the sensing element and the sound source to be collected different.

上述的声音信号采集处理奉统,其中,所述声音传感元件同轴同向设置, 且相邻的声音传感元件间隔预,距离。 The above-described sound signal acquisition and processing system instructions, wherein the sound sensing element is disposed coaxially in the same direction, and the acoustic sensing elements spaced adjacent pre-distance.

上述的声音信号采集处理系统,其中,所述声音信号预处理模块具体包 The above-described audio signal acquisition and processing system, wherein the sound signal pre-processing module is packet

括:差分放大器,与声音传感元件对应相连; Comprising: a differential amplifier corresponding to the sound sensing element is connected;

A/D转换模块,用于将差分放大器输出的信号转换为数字语音信号后输出给波束形成模块。 A / D conversion module for converting a signal output from the differential amplifier to the digital voice signal to the beam forming module.

上述的声音信号采集处理系统,其中,所述波束形成模块具体用于采用基于频域最小均方的自适应波束形成算法计算波束。 The above-described audio signal acquisition and processing system, wherein, the beam forming module is used to calculate the beamforming algorithm using frequency domain adaptive beam forming based on a minimum mean square.

上述的声音信号采集处理系统,其中,还包括: The above-described audio signal acquisition and processing system, wherein, further comprising:

一校正模块,设置于声音信号预处理模块和波束形成模块之间,用于补偿声音传感元件之间电气性能差异。 A correction module, the sound signal pre-processing module is provided, and between the beam-forming means for compensating the difference between the electrical properties of acoustic sensing elements.

上述的声音信号采集处理系统,其中,还包括: The above-described audio signal acquisition and processing system, wherein, further comprising:

一切换模块,用于选择一个声音传感元件输出的信号发送给声音处理芯片,或选择将所有声音传感元件的输出信号发送给对应的声音信号预处理模块。 A signal switching means for selecting one acoustic sensing element output is sent to the audio processing chip, or select all the transmission output signal of the acoustic sensing element corresponding to a sound signal pre-processing module.

为了更好的实现上述目的,l本发明还提供了一种声音信号釆集处理方法, 拾音面与需要采集的声源的距离不同的多个声音传感元件采集声音信号后, 经A/D转化、波束成形后,获取该波束内与第一频谱分量频率相同的部分, 并将该波束内与第一频谱分量麵率相同的部分进行D/A转换后输出,所述第一频谱分量在声音传感元件拾音面的声压差处于声压差预定值范围。 In order to achieve the above object, l The present invention also provides a method of processing a sound signal preclude set, different from the plurality of acoustic sensing elements collected sound signal from the sound source of the pickup surface to be collected, by the A / D conversion, after beamforming, obtaining the first spectral components within the same frequency of the beam portion, and after D / a conversion output of the same component surface portion of the beam within a first frequency spectrum, said first spectral component in the acoustic sensing element in sound pressure pickup surface is a predetermined sound pressure difference range. 上述的声音信号采集处理方法,其中,具体包括: 步骤S1,拾音面与需要采集的声源的距离不同的多个声音传感元件采集声音信号; , Collecting the above-described audio signal processing method, which comprises: a different distance step S1, the pickup surface of the sound source to be captured a plurality of acoustic sensing elements collect a sound signal;,

步骤S2,分别将多路采集P的声音信号转换为数字语音信号; 步骤S3,将所述多路数字l吾音信号利用波束形成算法获取波束; 步骤S4,获取所述波束内i与第一频谱分量频率相同的部分,并进行D/A 转换后输出。 Step S2, respectively, the multi-channel sound collecting signal into a digital voice signal P; step S3, the multi-channel digital audio signal I l utilizes beamforming to obtain beamforming algorithm; step S4, and acquiring said first beam i spectral component frequency outputs the same parts, and D / a conversion.

上述的声音信号采集处理好法,其中,步骤S4具体包括: The above-described method to handle the collected sound signal, wherein the step S4 further comprises:

步骤S41,获取声音传感无件的输出信号; Step S41, the acquired output signal without the acoustic sensing element;

步骤S42,计算输出信号的频谱分量在声音传感元件上的声压差; Step S42, the calculated spectral components of the output signal on the sound pressure of the acoustic sensing element;

步骤S43,获取在声音传感元件上的声压差处于声压差预定值范围的频谱步骤S44,将所述波束内与第一频谱分量频率相同的部分取出后进行D/A Step S43, the acquired sound on a sound pressure difference sensor element in a predetermined range of the sound pressure spectrum of step S44, the spectral components with the same frequency as the first portion of said beam is D / A after extraction

转换,并输出;或 Conversion, and outputs; or

步骤S44',将所述波束内与第一频谱分量频率不相同的部分滤除,并将 Step S44 ', the inner beam of the first frequency spectral components are not the same portion was filtered off, and

剩余部分进行D/A转换后输出; After the remaining portion of the D / A conversion output;

上述的声音信号采集处理方法,其中,所述步骤S1和步骤S2之间还包括: 步骤S5,将多路采集到的声音信号分别进行差分放大处理。 Collecting the above-described audio signal processing method, wherein, between the steps S1 and S2 further comprising: a step S5, the multi-channel sound signals collected by differentially amplifying process. 上述的声音信号采集处理方法,其中,所述波束形成算法为基于频域的 Collecting the above-described audio signal processing method, wherein, the beam forming algorithm is a frequency domain based

最小均方自适应波束形成算法。 Least Mean Square adaptive beamforming algorithms. ,

上述的声音信号采集处理方法,其中,所述步骤S1之前还包括: 步骤S6,判断是否需要启劝噪音消除处理,如果是进入步骤S1,否则选 Collecting the above-described audio signal processing method, wherein, before the step S1 further comprising: a step S6, it is determined whether to enable the noise-canceling advised process proceeds to step S1 if it is, or is selected from

择一个声音传感元件输出的信,发送到声音处理芯片。 Select a channel acoustic sensing element output, transmitted to the sound processing chip.

本发明的声音信号采集处*系统及方法,通过将声音传感元件同轴同向, * The sound collection signal systems and methods of the present invention, the acoustic sensing element is coaxial with the direction,

且间隔一定距离设置,这样在缺束形成后,将波束内声压差不处于预定值范 And arranged at a distance, so that the lack of beam forming, the pressure within the acoustic beam is not in a predetermined value range

围的频谱分量进行滤除,有效举消除了波束内和波束外的噪音的影响,提高 Wai filter out spectral components, effectively eliminating the effects of noise held within the outer beam and the beam, to improve the

了信号处理的质量。 The quality of the signal processing.

附图说明 BRIEF DESCRIPTION

图1为采用并排方式设置的麦克风阵列的示意图; 图2为本发明的实施例中声音信号采集处理系统的结构示意图; 图3为本发明的声音信号采集模块的声音传感元件的另外一种设置方式示意图; FIG 1 is a schematic diagram of the microphone array disposed parallel employed; FIG. 2 is a schematic structure of the collected sound signal processing system of the present embodiment of the invention; acoustic sensing element the sound signal acquisition module of the present invention. FIG. 3 is another schematic arrangement;

图4为本发明的实施例中声音信号采集处理系统中声音信号预处理模块的结构示意图; Schematic sound signal preprocessing module structure of the collected sound signal processing system according to the present invention. FIG. 4;

图5为本发明的实施例中声音信号采集处理系统的具体结构示意图; 图6为本发明的实施例中声音信号采集处理方法的流程示意图。 Figure 5 a schematic view of a specific configuration of the audio signal acquisition and processing system of the present embodiment of the invention; Examples schematic flow collected sound signal processing method of the embodiment of the present invention. FIG. 6.

具体实施方式 Detailed ways

下面结合附图对本发明进行详细的说明。 The following figures present invention will be described in detail in conjunction. 本发明的声音信号采集处瞎系统如图2所示,包括: A sound signal acquisition system of the present invention at the blind shown in Figure 2, comprising:

声音信号采集模块21,用于釆集声音信号,包括多个声音传感元件,其拾音面与需要采集的声源的距离不同,使得声源发出的声音信号在不同的声 Different acoustic sound signal a sound signal acquisition module 21, a set of sound signals Bian, comprising a plurality of acoustic sensing elements which pickup surface to be collected from the sound source is different from the sound source in such

7音传感元件上有不同的声压; There are 7 different sound components of the sound pressure;

声音信号预处理模块22,与声音信号采集模块数目连接,用于将各声音传感元件采集到的声音信号分别转换成数字语音信号;该声音信号预处理模块22分别对多路采集的声音信号进行转换处理,并分别输出到波束形成模块23; 22, the number of the sound signal acquisition module and a sound signal pre-processing module is connected to each of the acoustic sensing element for collecting sound signals are converted into digital voice signals; the audio signal preprocessing module 22 are the multiplexed audio signal acquisition conversion processing, and outputs to the beam forming module 23, respectively;

波束形成模块23,用于根据转换后的多路数字语音信号,并利用基于频域LMS (Least Mean Square,最小均方)的自适应波束形成算法获取波束, 滤除波束外的噪音信号; A beam forming module 23, according to multi-channel digital speech signal conversion, and is formed by using a frequency domain algorithm based on LMS (Least Mean Square, Least Mean Square) adaptive beam acquisition beam filtered noise signals outside the beam;

噪音消除模块24,用于获取波束内与第一频谱分量频率相同的部分并发送,该第一频谱分量在声音传感元件上的声压差处于声压差预定值范围; Noise cancellation module 24, configured to obtain the frequency spectral components of the first beam portions and transmitting the same, the first spectral components in the sound pressure of the acoustic sensing element is in a predetermined value range sound pressure difference;

D/A转换模块25,用于将噪音消除模块24输出的信号进行D/A转换,并发送给相应的声音处理芯片进行处理。 D / A converter module 25, for noise cancellation signal output module 24 performs D / A conversion, and sends a corresponding audio processing chips for processing.

在此,声音信号采集模块中的声音传感元件可釆用同轴同向设置,且相邻的声音传感元件之间间隔预定的距离Ar,这样,声源发出的声音信号在不同的声音传感元件上就有不同的声压(这点将在后面进行详细的描述)。 Here, the sound signal acquisition module may preclude the use of acoustic sensing element disposed coaxially to the same, and the interval between adjacent predetermined acoustic sensing element from Ar, so that the sound signal from the sound source in different voices on the sensor elements have different sound pressure (which will be described in detail later).

其中声音传感元件都包括一个拾音面,在此,同向设置是指声音传感元件的拾音面朝向相同的方向。 Wherein acoustic sensing element comprises a pickup surface, in this case, the same setting is toward the same surface of the acoustic sensing element of the pickup direction.

在此声音传感元件同轴同向是本发明最佳的一种方式,如图3所示,如果声音传感元件的拾音面之间的夹角小于门限夹角,声音传感元件的拾音面的轴线之间的距离小于门限轴线距离也是允许的。 The acoustic sensing element is coaxial with the best aspect of the present invention, as shown, if the angle between the pickup surface of the acoustic sensing element 3 is less than the threshold angle, the acoustic sensing element the distance between the axes of the pickup surface is smaller than a threshold distance from the axis is also allowed.

本发明的具体实施例中,以同轴同向,相邻声音传感元件之间间隔预定的距离来进行说明,但如图3凍示的其他的声音传感元件设置方式也是可以的,只要能使得声源发出的声音信号在声音传感元件上表现出不同的声压就可以。 Specific embodiments of the present invention, coaxial to the same direction, the spacing between adjacent elements of the acoustic sensing a predetermined distance to be described, but three freeze FIG other acoustic sensing element arrangement shown are possible, as long as enables the sound signal from the sound source to the sound performance of different sensor elements can be sound pressure.

如图2所示,声音信号采紫漠块21,包括多个同轴同向设置的声音传感元件,且相邻的声音传感元件之间间隔预定的距离Ar。 As shown, a predetermined interval between the sound signal acquisition block 21 desert Violet, comprising a plurality of coaxially disposed with the sound sensing element, the sensing element and the adjacent sound distance Ar 2.

通常距声源愈远的点声强k小,若不考虑介质对声能的吸收,点声源在 Strong typically farther away from the sound source point sound small k, without considering the sound absorption energy medium, the point sound source

自由声场中向四周均匀辐射声能时,距声源r处的声强为"『/4一,其中I为距点声源距离为r处的声强,W为点声源功率。若s表示包围声源的封闭面面积,声功率w和声强i的关系为:『=《4^s, Free field is enabled to the surrounding uniformly radiated sound, sound from a sound source at r intensity of "" / 4 a, wherein I is from point sound source distance is sound at r strength, W is a point sound source power. If s represents a sealing surface surrounding the area of ​​the sound source, sound intensity relationship between acoustic power w i is: "=" 4 ^ s,

其中/"是声强在微元面积dS法线方向的分量。 Where / "is the sound intensity component infinitesimal area dS normal direction.

而在自由与半自由声场中,点声源的声压级与声功率级的关系式分别为: W201gr-11 卿i,;-201gr-8 卿 In the free half free field, the relationship of the sound pressure and sound power levels are point sound source: W201gr-11 Qing i,; - 201gr-8 Qing

若测点l, 2与声源的距离分别为n、 r2,则由n至r2的声压级衰减量为: If the measured point l, 2 distance from the sound source n, r2, r2 by n to a sound pressure level attenuation amounts were as follows:

距离每增加一倍,声压级衰减大约6dB。 Each doubling of the distance, the sound pressure level attenuation of about 6dB.

AL尸二丄尸,一丄P2-201gi (dB)。 AL Shang two dead corpse, a Shang P2-201gi (dB).

本发明中,利用同轴同心设置的麦克风模组,"=rl + Ar,在此假设Ar为 In the present invention, with the microphone module arranged concentrically coaxially, "= rl + Ar, Ar is assumed here

5mm,则: 5mm, then:

Alp = 201g(l + A = 201g(l+丄)(dB ) r, ,, Alp = 201g (l + A = 201g (l + Shang) (dB) r, ,,

由上式可以得到以下对应关系(在此,仅列举了各数量级下的对应关系, 以得出一个结论): The following formula can be obtained by the correspondence relation (in this case, only lists the correspondence between the various orders of magnitude, to arrive at a conclusion):

n为10mm,两个声音传感元件上的声压差为3.521825dB; n is 10mm, a sound pressure difference over the two acoustic sensing elements 3.521825dB;

n为50mm,两个声音传感元件上的声压差为0.827854dB; n is 50mm, a sound pressure difference on the two acoustic sensing elements 0.827854dB;

n为100mm,两个声音传感元件上的声压差为0.423786dB; n is 100mm, sound pressure difference on the two acoustic sensing elements 0.423786dB;

^为300mm,两个声音传寧元件上的声压差为0.143572dB; ^ Is 300mm, two voice sound pressure difference on the element rather 0.143572dB;

n为500mm,两个声音传感元件上的声压差为0.086427dB; n is 500mm, sound pressure difference on the two acoustic sensing elements 0.086427dB;

n为1500mm,两个声音传感元件上的声压差为0.028985dB。 n is 1500mm, sound pressure difference on the two acoustic sensing elements 0.028985dB.

从以上的对应关系可以看出,当n较大(如500mm以上)时,发声点在两个麦克风上形成的声压差已经相当小(如500mm以上,形成的声压差小于O.ldB)。 As can be seen from the above correspondence relationship, when n is large (e.g., 500mm or more), the utterance sound pressure difference dot formed on the two microphones have a relatively small (e.g., 500mm or more, the sound pressure differential is less than O.ldB) .

但一般来讲,声音信号采te模块距离其需要采集的声源的距离都不会太 But generally speaking, the sound signal acquisition module te its distance from the sound source to be collected will not be too

大,如手机中一般不会超过200mm,麦克风中也一般不会超过300mm,因此, 其声压差(通过转换后为电压)基本都可以有效辨别。 Large, such as mobile phones generally not more than 200mm, the microphone is also generally not more than 300mm, thus, the sound pressure (converted into a voltage by) basically can effectively distinguish.

因此,在噪音消除模块中i完全可以将波束内声压差处于声压差预定值范围之外的频谱分量滤除(或者仅提取声压差处于声压差预定值范围的频谱分量),从而将波束内的噪音滤除,获得更干净的信号。 Accordingly, in the noise-canceling module i can be the sound pressure beam in spectral components outside the predetermined value range of the sound pressure filtered (or sound pressure difference to extract only the sound pressure difference of spectral components in a predetermined value range), such that the noise in the filtered beam to obtain a cleaner signal. 其中,声音信号采集模块中的声音传感元件的距离可根据实际的应用进 Wherein acoustic sensing element from the sound signal acquisition module according to the actual application may be into

行设置,如r,较大时,可将声^传感元件的距离设置得稍大些,反之就可将声音传感元件的距离设置稍小些,当然,也可以通过在噪音消除模块24设置一声压差预定值保存模块,用于修改并保存声压差预定值,噪音消除模块24根据保存的声压差预定值进行噪音的滤除。 Line sets, such as r, is large, the distance may be ^ acoustic sensing element is set to be slightly larger, whereas the distance of the acoustic sensing element can be set slightly smaller, of course, can also in the noise-canceling module 24 disposing an acoustic pressure predetermined value storage module configured to modify and save sound pressure difference predetermined value, the noise cancellation module 24 for filtering out noise sound pressure difference according to a predetermined stored value. 如需要被采集的声源距离较远时, 可将声压差预定值设置小些,这样就可以保证需要的声音被采集(或者说不被滤除),反之就可将该声压差预定值设置稍大些。 When far as, the predetermined value may be set sound pressure difference from the sound source to be captured smaller, so as to ensure the desired sound to be acquired (or say is filtered), whereas the difference in sound pressure can be predetermined value slightly larger. 当然,也可利用该声压差预定值保存模块保存多个声压差预定值,由用户根据采集的声源距离来选择对应的声压差预定值,保证需要的声音被采集,而不需要的声音(噪音) 被滤除。 Of course, this may also be utilized in sound pressure value holding module stores a plurality of predetermined sound pressure difference a predetermined value, the user selects a sound pressure difference by the predetermined value corresponding to the collected sound source distance, to ensure the desired sound to be collected, without sound (noise) are filtered out.

同时,在声音信号釆集模块走线较长时,可能会带来共模噪声,因此, 如图4所示,本发明的声音信号预处理模块具体包括: 差分放大器,用于过滤、rt制共模噪声; Meanwhile, when the sound signal is set longer preclude the trace module may be of common mode noise, and therefore, as shown in Figure 4, the sound signal pre-processing module of the present invention comprises: a differential amplifier for filtering, rt system common mode noise;

A/D转换模块,用于将差分放大器输出的信号转换为数字语音信号。 A / D converting means for converting the signal output of the differential amplifier into a digital voice signal. 波束形成模块采用基于频域LMS的自适应波束形成算法计算波束,滤除 It calculates beam forming module is formed using a frequency domain LMS algorithm based on an adaptive beamforming beams filtered

波束外的噪音信号。 Noise signals outside the beam.

同时,由于本发明采用多个声音传感元件的采集信号进行处理,因此为 Meanwhile, since the present invention employs a plurality of acoustic sensing elements pickup signal is processed, therefore

了达到更好的效果,本发明的声音信号采集处理系统还在声音信号预处理模 To achieve a better results, the collected sound signal processing system of the present invention is also a sound signal pre-mold

块和波束形成模块之间设置一校正模块,用于补偿声音传感元件之间电气性 It is provided between the block and the beam-forming module a correction means for compensating the electrical resistance between the acoustic sensing element

能差异。 Energy difference.

同时,考虑到将声音信号来集处理系统应用到任何场合都需要消耗能量, 但在背景噪音不大的情况下,不需要对背景噪音过滤即可实现清楚地语音还原,此时为了降低功耗、延长使用时间,此时可采用一个声音传感元件的信号即可,因此本发明的声音信号采集处理系统还包括: Meanwhile, considering the sound signal processing system set applied to the energy must be consumed on any occasion, but not in the background noise, the background noise is not necessary to achieve a clear voice to the filter reduced in order to reduce the power consumption at this time , extended periods of time, this time can be a sound signal to the sensing element, the collected sound signal processing system of the present invention further comprises:

切换模块,用于选择一个声音传感元件输出的信号,并发送到声音处理芯片(如手机的基带麦克风输入电路)进行后续处理,或用于将所有声音传感元件的输出信号发送给声音信号预处理模块进行差分放大、A/D转换等操作。 Signal switching means for selecting the output of a sound sensing element, and sent to the audio processing chips (e.g., baseband phone microphone input circuit) for subsequent processing, or for transmitting an output signal of the sensing element of all the sound signals to the sound differentially amplifying the preprocessing module, A / D conversion operation.

该切换模块的切换操作可由用户通过按键、软件图标或菜单选项等进行切换,由于其实现为本领域普通技术人员所熟知,在此不再赘述。 Switching operation of the switching module by a user through a key, the software icon or menu option to switch the like, since its implementation by those of ordinary skill in the art, are not repeated here. 噪音消除模块具体执行以下操作: Noise cancellation module is the following:

首先获取声音传感元件的输出信号; 计算对应频谱分量在声音传感元件上的声压差; First acquires an output signal of the acoustic sensing element; calculates the corresponding spectral components in the sound pressure of the acoustic sensing element;

获取在声音传感元件上的声压差在声压差预定值范围的频谱分量; Acoustic sensing element acquires the sound pressure in the sound pressure value of a predetermined range of spectral components;

根据声音传感元件上的声压差在声压差预定值范围的频谱分量对应获取波束内的同频信号。 The acoustic sound pressure on the sensing element corresponding to the sound pressure value of a predetermined range of spectral components of signals in the same frequency acquisition beam.

最后,本发明的声音信号釆集处理系统的一个最佳实施例的结构图如图5 Finally, a block diagram of the preferred embodiment of the present invention, the sound collection signal processing system of FIG Bian 5

所示,图中为釆用两个声音传感元件的声音信号采集处理系统。 As shown in the figure is preclude sound signal acquisition and processing system with two acoustic sensing element. 下面结合附图对本发明的声音信号采集处理方法进行进一步说明。 The following drawings collected sound signal processing method of the present invention will be further described in conjunction.

如图6所示,本发明的声音信号采集处理方法包括如下步骤-步骤61,多个同轴同向设匿,且相邻的声音传感元件之间间隔预定的距离Ar的声音传感元件采集外界声音信号; As shown, the collected sound signal processing method of the present invention comprises the steps 6 - Step 61, with a plurality of coaxially arranged to hide, the acoustic sensing element spaced a predetermined distance between Ar and adjacent the acoustic sensing element capturing ambient sound signal;

步骤62,声音信号预处理模块对应将声音信号采集模块釆集到的声音信 Step 62, a sound signal pre-processing module corresponding to the sound signal acquisition module set to preclude the audio signal

号转换成数字语音信号; No. converted into a digital voice signal;

步骤63,波束形成模块根据转换后的数字语音信号利用基于频域LMS (Least Mean Square,最小均方)的自适应波束形成算法计算波束,滤除波束外的噪音信号; Step 63, beam forming algorithm on the frequency domain beam forming module LMS (Least Mean Square, Least Mean Square) adaptive beam in accordance with the converted digital voice signal using the filtered noise signals outside the beam;

步骤64,噪音消除模块提取波束内声压差处于声压差预定值范围的频谱分量后输出; Step 64, after the noise cancellation module extracts the sound pressure beam in a predetermined sound pressure difference of spectral components of the output value range;

步骤65, D/A转换模块将噪音消除模块输出的信号进行D/A转换后输出。 Step 65, D / A conversion module noise cancellation signal outputted from the module after D / A conversion output. 步骤64具体包括: 64 comprises the step of:

步骤641,噪音消除模块获取声音传感元件的输出信号; 步骤642,噪音消除模块计算对应频谱分量在声音传感元件上的声压差; 步骤643,噪音消除模块获取在声音传感元件上的声压差在声压差预定值范围的频谱分量; Step 641, the noise cancellation output signal of the acoustic sensing element acquisition module; step 642, the noise cancellation module calculates the corresponding spectral components in the sound pressure of the acoustic sensing element; step 643, in the noise cancellation module acquires the acoustic sensing element sound pressure difference spectral components in the sound pressure of a predetermined value range;

步骤644,噪音消除模块竭据声音传感元件上的声压差在声压差预定值范 Step 644, the noise cancellation module exhaust sound pressure difference of the acoustic sensing element according to the sound pressure at a predetermined value range

围的频谱分量对应获取波束内!的同频信号后输出。 Outputs the spectral components around the same frequency corresponding to the acquired signal beam! A. 同时,步骤61和步骤62之间还包括:步骤66,差分放大器对声音传感元件的输出信号进行差分放大处理,消 Meanwhile, between steps 61 and 62 further comprising the step of: step 66, the differential amplifier output signal of the acoustic sensing element differential amplification processing, eliminate

除共模噪音。 In addition to common mode noise.

同时,本发明步骤62和步骤63之间还包括: Meanwhile, the present invention between step 62 and 63 further comprising the step of:

步骤67,校正模块补偿声音传感元件之间电气性能差异,其具体执行以下操作: Step 67, the electrical performance difference between the correction module compensates for the acoustic sensing element, which carries out the following operations:

对每个麦克风分别进行等声压、全频段扫描输入,分别获取各自的频谱及采样音量。 Separately for each microphone sound pressure and the like, the whole band scan input, respectively acquire each spectrum and sample volume. 对比麦克风规格书,对各自的差异之处进行补偿。 Comparative microphone specification, each of the differences is compensated. 从而实现麦克风之间的差异性校正。 Thereby realizing the correction difference between the microphones.

同时本发明步骤61之前还包括: Meanwhile prior to the step 61 of the present invention further comprises:

步骤68,判断是否需要启勒噪音消除处理,如果是进入步骤61,否则选择一个声音传感元件输出的信号发送到声音处理芯片(如手机的基带麦克风输入电路)。 Step 68, determine whether to enable Le noise-canceling processing, if proceeds to step 61, otherwise a selection signal output from the acoustic sensing element is sent to the audio processing chips (e.g., mobile phones with a microphone input circuit group).

以上所述仅是本发明的优选实施方式,应当指出,对于本技术领域的普通技术人员来说,在不脱离本发明原理的前提下,还可以作出若干改进和润饰,这些改进和润饰也应视为本发明的保护范围。 The above are only preferred embodiments of the present invention, it should be noted that those of ordinary skill in the art, in the present invention without departing from the principles of the premise, further improvements and modifications may be made, these improvements and modifications should also be the protection scope of the present invention.

Claims (12)

1. 一种声音信号采集处理系统,包括声音信号采集模块、声音信号预处理模块、波束形成模块和D/A转换模块,其特征在于,还包括:噪音消除模块,用于获取波束形成模块输出的波束中频率与第一频谱分量频率相同的部分,并发送给D/A转换模块,所述第一频谱分量为声音传感元件的输出信号的对应频谱分量中,在声音传感元件拾音面的声压差处于声压差预定值范围的所有频谱分量;所述声音信号采集模块包括多个声音传感元件,不同声音传感元件的拾音面与需要采集的声源的距离不同。 1. A sound signal acquisition and processing system, comprising a sound signal acquisition module, a sound signal pre-processing module, a beam forming module and a D / A conversion module, which is characterized in that, further comprising: a noise cancellation module, configured to obtain an output beam forming module of the same frequency as the beam portion of the first component of the frequency spectrum, and sent to D / a converter module, the corresponding output signal of the spectral components of first spectral component of the acoustic sensing element, the sensing element in the sound pickup sound pressure difference in the plane of all spectral components of the sound pressure difference a predetermined value range; the sound signal acquisition module comprises a plurality of acoustic sensing elements, different distances from the sound source of the pickup surface to be collected different acoustic sensing element.
2. 根据权利要求l所述的声音信号采集处理系统,其特征在于,所述声音传感元件同轴同向设置,且相邻的声音传感元件间隔预定距离。 The collected sound signal processing system of claim l, wherein said acoustic sensing element disposed coaxially in the same direction, and adjacent the acoustic sensing element spaced a predetermined distance.
3. 根据权利要求1所述的声音信号采集处理系统,其特征在于,所述声音信号预处理模块具体包括:差分放大器,与声音传感元件对应相连;A/D转换模块,用于将差分放大器输出的信号转换为数字语音信号后输出给波束形成模块。 3. The sound signal acquisition and processing system according to claim 1, characterized in that, the sound signal pre-processing module comprises: a differential amplifier connected to the sensor element corresponding to the sound; A / D conversion means for differential amplifier output signal is converted to a digital speech signal to the beam forming module.
4. 根据权利要求1所述的声音信号采集处理系统,其特征在于,所述波束形成模块采用基于频域最小均方的自适应波束形成算法。 4. The collected sound signal processing system of claim 1, wherein said beam forming module is formed using frequency domain adaptive beamforming algorithm based on minimum mean square.
5. 根据权利要求1所述的声音信号采集处理系统,其特征在于,还包括: 一校正模块,设置于声音信号预处理模块和波束形成模块之间,用于补偿声音传感元件之间电气性能差异。 The collected sound signal processing system according to claim 1, characterized in that, further comprising: a correction module, a sound signal is provided to a pre-processing module between the module and the beam forming to compensate for electrical connection between the acoustic sensing element performance differences.
6. 根据权利要求l、 2、 3、 4或5所述的声音信号采集处理系统,其特征在于,还包括:一切换模块,用于选择一个声音传感元件输出的信号发送给声音处理芯片,或选择将所有声音传感元件的输出信号发送给对应的声音信号预处理模块。 According to claim l, 2, the sound signal acquisition and processing system 3, 4 or 5, characterized in that, further comprising: a switching means for selecting one of the acoustic sensing element output signal is sent to the voice processing chip or select all of the transmission output signal of the acoustic sensing element corresponding to a sound signal pre-processing module.
7. —种声音信号采集处理方法,其特征在于,拾音面与需要采集的声源的距离不同的多个声音传感元件采集声音信号后,经A/D转化、波束成形后, 获取该波束内与第一频谱分量频率相同的部分,并将该波束内与第一频谱分量频率相同的部分进行D/A转换后输出,所述第一频谱分量为声音传感元件的输出信号的对应频谱分量中,在声音传感元件拾音面的声压差处于声压差预定值范围的所有频谱分量。 7. - species collected sound signal processing method, wherein, after the sound collection signal, the A / D conversion, beamforming pickup surface at different distances from the sound source to be captured after the plurality of acoustic sensing elements, to obtain the after, and the D / a-converted beams within the same component part of the first frequency spectrum in the beam of the same portion of the frequency spectral components of the first output, the first spectral component of an output signal corresponding to the sound sensing element spectral components, the sound pressure difference of the acoustic sensing element pickup plane of all spectral components in the sound pressure difference of the predetermined value range.
8. 根据权利要求7所述的声音信号采集处理方法,其特征在于,具体包括:步骤S1,拾音面与需要采集的声源的距离不同的多个声音传感元件采集声音信号;步骤S2,分别将多路采集到的声音信号转换为数字语音信号; 步骤S3,将所述多路数字语音信号利用波束形成算法获取波束; 歩骤S4,获取所述波束内与第一频谱分量频率相同的部分,并进行D/A 转换后输出。 8. The collected sound signal processing method according to claim 7, characterized in that comprises: a step S1, the pickup surfaces at different distances from the sound source to be captured a plurality of acoustic sensing elements pick up sound signal; step S2 , respectively, to the multi-channel sound collecting signal into a digital voice signal; step S3, the multi-channel digital voice signal utilizes beamforming to obtain beamforming algorithm; ho step S4, the beam acquiring the same as the first spectral component frequency the output section, and D / a converted.
9. 根据权利要求8所述的声音信号采集处理方法,其特征在于,步骤S4具体包括:步骤S41,获取声音传感元件的输出信号;歩骤S42,计算输出信号的频谱分量在声音传感元件上的声压差;步骤S43,获取输出信号中在声音传感元件上的声压差处于声压差预定值范围的频谱分量;步骤S44,将所述波束内与第一频谱分量频率相同的部分取出后进行D/A 转换,并输出;或步骤S44',将所述波束内与第一频谱分量频率不相同的部分滤除,并将剩余部分进行D/A转换后输出。 9. The collected sound signal processing method according to claim 8, wherein the step S4 further comprises: step S41, the acquired output signal of the acoustic sensing element; ho step S42, the spectral components of the output signal is calculated in the sound sensor acoustic pressure differential across the element; step S43, the acquired output signal of the acoustic sound pressure on the sensing element at a predetermined value of spectral components in sound pressure range; step S44, the same beam within the first spectral component frequency fractions were taken after D / a conversion, and outputs; or step S44 ', the first spectral components with the same frequency is not filtered out within the beam portion, and the remaining portion after D / a conversion output.
10. 根据权利要求8或9所述的声音信号采集处理方法,其特征在于,所述步骤S1和步骤S2之间还包括:步骤S5,将多路采集到的声音信号分别进行差分放大处理。 10. The collected sound signal processing method of claim 8 or claim 9, characterized in that, between the steps S1 and S2 further comprising: a step S5, the multi-channel sound signals collected by differentially amplifying process.
11. 根据权利要求8或9所述的声音信号采集处理方法,其特征在于,所述波朿形成算法为基于频域的最小均方自适应波束形成算法。 11. The collected sound signal processing method of claim 8 or claim 9, wherein said algorithm is a wave form formed Bouquet minimum mean square adaptive algorithm based on the beam in the frequency domain.
12. 根据权利要求8或9所述的声音信号采集处理方法,其特征在于,所述歩骤S1之前还包括:步骤S6,判断是否需要启动噪音消除处理,如果是进入歩骤S1,否则选择一个声音传感元件输出的信号发送到声音处理芯片。 12. The collected sound signal processing method of claim 8 or claim 9, characterized in that, prior to the step S1 ho further comprising: a step S6, it is determined whether to start the noise canceling processing, if ho step S1 is entered, otherwise select a voice sensing element signal output from the transmitter to the sound processing chip.
CN 200610151451 2006-09-08 2006-09-08 Sound signal collecting and processing system and method thereof CN100505041C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 200610151451 CN100505041C (en) 2006-09-08 2006-09-08 Sound signal collecting and processing system and method thereof

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 200610151451 CN100505041C (en) 2006-09-08 2006-09-08 Sound signal collecting and processing system and method thereof

Publications (2)

Publication Number Publication Date
CN101140760A CN101140760A (en) 2008-03-12
CN100505041C true CN100505041C (en) 2009-06-24

Family

ID=39192681

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 200610151451 CN100505041C (en) 2006-09-08 2006-09-08 Sound signal collecting and processing system and method thereof

Country Status (1)

Country Link
CN (1) CN100505041C (en)

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5168162B2 (en) * 2009-01-16 2013-03-21 沖電気工業株式会社 Sound signal adjustment device, program and method, and telephone device
CN102376309B (en) * 2010-08-17 2013-12-04 骅讯电子企业股份有限公司 System and method for reducing environmental noise as well as device applying system
KR101782050B1 (en) * 2010-09-17 2017-09-28 삼성전자주식회사 Apparatus and method for enhancing audio quality using non-uniform configuration of microphones
CN102981149B (en) * 2012-11-27 2014-07-30 无锡儒安科技有限公司 Down-hole remote locating method and system based on beam forming
CN104036783B (en) * 2014-05-19 2017-07-18 孙国华 The magnetic resonance imaging apparatus scan adaptive speech enhancement system
CN105100413B (en) * 2015-05-27 2018-08-07 努比亚技术有限公司 Information processing method and apparatus, the terminal
CN104994456B (en) * 2015-07-07 2019-07-09 惠州Tcl移动通信有限公司 A kind of earphone and its method improving call tone quality
CN105049802B (en) * 2015-07-13 2018-06-19 深圳警翼智能科技股份有限公司 Kind of speech recognition and enforcement recorder identification method
CN108174320A (en) * 2018-01-29 2018-06-15 广东思派康电子科技有限公司 Adaptive ear canal active noise reduction earphone and adaptive ear canal active noise reduction method

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2000132200A (en) 1998-10-27 2000-05-12 Matsushita Electric Ind Co Ltd Audio/video device with voice recognizing function and voice recognizing method
CN1296258A (en) 1999-11-10 2001-05-23 三菱电机株式会社 Noise canceller
CN1416564A (en) 2000-01-12 2003-05-07 声波创新有限公司 Noise reduction appts. and method
JP2004271596A (en) 2003-03-05 2004-09-30 Advanced Media Inc Speech recognition system
CN1624762A (en) 2003-12-02 2005-06-08 雅马哈发动机株式会社 Helmet incorporating an active noise cancellation system, vehicle system employing same, and method of canceling noise in a helmet
CN1670823A (en) 2004-03-17 2005-09-21 哈曼贝克自动系统股份有限公司 Method for detecting and reducing noise from a microphone array

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2000132200A (en) 1998-10-27 2000-05-12 Matsushita Electric Ind Co Ltd Audio/video device with voice recognizing function and voice recognizing method
CN1296258A (en) 1999-11-10 2001-05-23 三菱电机株式会社 Noise canceller
CN1416564A (en) 2000-01-12 2003-05-07 声波创新有限公司 Noise reduction appts. and method
JP2004271596A (en) 2003-03-05 2004-09-30 Advanced Media Inc Speech recognition system
CN1624762A (en) 2003-12-02 2005-06-08 雅马哈发动机株式会社 Helmet incorporating an active noise cancellation system, vehicle system employing same, and method of canceling noise in a helmet
CN1670823A (en) 2004-03-17 2005-09-21 哈曼贝克自动系统股份有限公司 Method for detecting and reducing noise from a microphone array

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
语音识别中的抗噪声方法. 崔屹,张东,时良平,陈丽媛.北京邮电大学学报,第21卷第2期. 1998

Also Published As

Publication number Publication date
CN101140760A (en) 2008-03-12

Similar Documents

Publication Publication Date Title
RU2456701C2 (en) Higher speech intelligibility with application of several microphones on several devices
US8954324B2 (en) Multiple microphone voice activity detector
KR101120970B1 (en) Automatic volume and dynamic range adjustment for mobile audio devices
CA2527461C (en) Reverberation estimation and suppression system
RU2483439C2 (en) Robust two microphone noise suppression system
CN102257560B (en) Active audio noise cancelling
EP1312162B1 (en) Voice enhancement system
EP1742509B1 (en) A system and method for eliminating feedback and noise in a hearing device
EP2701145B1 (en) Noise estimation for use with noise reduction and echo cancellation in personal communication
CN1535555B (en) Acoustic devices, system and method for cardioid beam with desired null
EP2353159B1 (en) Audio source proximity estimation using sensor array for noise reduction
AU771444B2 (en) Noise reduction apparatus and method
JP4731115B2 (en) Improvement of speech intelligibility using psychoacoustic model and oversampled filter bank
CN100593351C (en) Array device and a portable terminal
US20060222184A1 (en) Multi-channel adaptive speech signal processing system with noise reduction
US20040193411A1 (en) System and apparatus for speech communication and speech recognition
JP5762956B2 (en) System and method for providing noise suppression utilizing nulling denoising
US20040185804A1 (en) Microphone device and audio player
US9460701B2 (en) Systems and methods for adaptive noise cancellation by biasing anti-noise level
CN105981408B (en) System and method for the secondary path information between moulding audio track
KR101463324B1 (en) Systems, methods, devices, apparatus, and computer program products for audio equalization
JP5678023B2 (en) Enhanced blind source separation algorithm for highly correlated mixing
CN104246870B (en) In the ear speaker adaptive noise cancellation channel (ANC) coordinated control of
CA2560034C (en) System for selectively extracting components of an audio input signal
US10382864B2 (en) Systems and methods for providing adaptive playback equalization in an audio device

Legal Events

Date Code Title Description
C06 Publication
C10 Entry into substantive examination
C14 Grant of patent or utility model
CF01