CN104952449A - Method and device for identifying environmental noise sources - Google Patents
Method and device for identifying environmental noise sources Download PDFInfo
- Publication number
- CN104952449A CN104952449A CN201510304135.1A CN201510304135A CN104952449A CN 104952449 A CN104952449 A CN 104952449A CN 201510304135 A CN201510304135 A CN 201510304135A CN 104952449 A CN104952449 A CN 104952449A
- Authority
- CN
- China
- Prior art keywords
- noise
- sound source
- feature
- information
- detected
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Landscapes
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
An embodiment of the invention provides a method and a device for identifying environmental noise sources. By the aid of the method and the device, problems of high computational complexity and serious impact on noise source identification efficiency of methods for identifying noise sources in the prior art can be solved. The method includes acquiring second features of to-be-detected noise by the aid of the device for identifying the environmental noise sources; carrying out vector quantization on the acquired second features according to preset rules to obtain to-be-matched information; comparing the to-be-matched information to sample information in standard feature sample databases, acquiring certain sample information from the standard feature sample databases and utilizing noise source types of known noise source types corresponding to the certain sample information as noise source types of the to-be-detected noise. Matching degrees of the certain sample information and the to-be-matched information reach preset threshold values. The method and the device have the advantages that environmental noise source identification computational complexity can be obviously lowered, the identification efficiency can be improved, and the method and the device are convenient to implement and easy to popularize and apply.
Description
Technical field
The present invention relates to Noise Recognition, in particular to a kind of neighbourhood noise identification of sound source method and device.
Background technology
Under Noise Sources Identification refers to the complicated sound source situation having many noise sources at the same time or comprise many vibration generation parts, in order to determine the performance of the sound radiation of each sound source or vibrating mass, distinguish noise source, and the measuring and analysis according to them, the effect of producing being graded and carry out.
In prior art, mainly based on the BP trained (Back Propagation) neural network, BP neural network corresponding for noise source to be identified and the BP neural network trained are contrasted, identify noise source, inventor finds after deliberation, this recognition method computation complexity is comparatively large, has had a strong impact on noise source recognition efficiency.
Summary of the invention
In view of this, the object of the embodiment of the present invention is to provide a kind of neighbourhood noise identification of sound source method and device, comparatively large to improve noise source recognition methods computation complexity in prior art, has had a strong impact on the problem of noise source recognition efficiency.
To achieve these goals, the technical scheme of embodiment of the present invention employing is as follows:
First aspect, embodiments provide a kind of neighbourhood noise identification of sound source method, be applied to neighbourhood noise identification of sound source device, standard feature Sample Storehouse is preset with in described neighbourhood noise identification of sound source device, described standard feature Sample Storehouse comprises and the fisrt feature of the noise of known sound source type is carried out by preset rules the sample information that vector quantization obtains, described sample information is corresponding with the noise of described known sound source type, and described method comprises:
Described neighbourhood noise identification of sound source device obtains the second feature of noise to be detected;
By described preset rules, the described second feature obtained is carried out vector quantization and obtain information to be matched;
Sample information in described information to be matched and described standard feature Sample Storehouse is contrasted, from described standard feature Sample Storehouse, obtain the sample information reaching predetermined threshold value with described information matches degree to be matched, using the sound source type of the sound source type of the noise of the known sound source type corresponding with the sample information that described information matches degree to be matched reaches predetermined threshold value as described noise to be detected.
In conjunction with first aspect, embodiments provide the first possible embodiment of first aspect, wherein, described fisrt feature and second feature are Mel frequency cepstral coefficient MFCC feature.
In conjunction with first aspect, embodiments provide the embodiment that the second of first aspect is possible, wherein, described neighbourhood noise identification of sound source device obtains the second feature of noise to be detected, comprising:
Described noise to be detected is carried out pre-emphasis, framing and adds Hamming window process;
For each frame signal of described noise to be detected, carry out Fourier transform respectively, obtain the amplitude spectrum of each frame signal described;
The second feature that logarithm operation and discrete cosine transform obtain described noise to be detected is done to described amplitude spectrum.
In conjunction with the embodiment that the second of first aspect is possible, embodiments provide the third possible embodiment of first aspect, wherein, described each frame signal for described noise to be detected, carry out Fourier transform respectively, after obtaining the amplitude spectrum of each frame signal described, also comprise:
Filtering is carried out to described amplitude spectrum;
Described the second feature that logarithm operation and discrete cosine transform obtain described noise to be detected is done to described amplitude spectrum, comprising:
The second feature that logarithm operation and discrete cosine transform obtain described noise to be detected is done to filtered described amplitude spectrum.
In conjunction with first aspect, or the first or the second of first aspect or the third possible embodiment, embodiments provide the 4th kind of possible embodiment of first aspect, wherein, described sample information in described information to be matched and described standard feature Sample Storehouse to be contrasted, from described standard feature Sample Storehouse, obtain the sample information reaching predetermined threshold value with described information matches degree to be matched, comprising:
Adopt Hidden Markov Model (HMM) the sample information in described information to be matched and described standard feature Sample Storehouse to be contrasted, from described standard feature Sample Storehouse, obtain the sample information reaching predetermined threshold value with described information matches degree to be matched.
Second aspect, embodiments provides a kind of neighbourhood noise identification of sound source device, comprising:
Preset unit, for preset standard feature samples storehouse, described standard feature Sample Storehouse comprises and the fisrt feature of the noise of known sound source type is carried out by preset rules the sample information that vector quantization obtains, and described sample information is corresponding with the noise of described known sound source type;
Obtain unit, for obtaining the second feature of noise to be detected;
Vector quantization unit, obtains information to be matched for the described second feature that described acquisition unit obtains being carried out vector quantization by described preset rules;
Choose unit, for the sample information in described information to be matched and described standard feature Sample Storehouse is contrasted, from described standard feature Sample Storehouse, obtain the sample information reaching predetermined threshold value with described information matches degree to be matched, using the sound source type of the sound source type of the noise of the known sound source type corresponding with the sample information that described information matches degree to be matched reaches predetermined threshold value as described noise to be detected.
In conjunction with second aspect, embodiments provide the first possible embodiment of second aspect, wherein, described fisrt feature and second feature are Mel frequency cepstral coefficient MFCC feature.
In conjunction with second aspect, embodiments provide the embodiment that the second of second aspect is possible, wherein, described acquisition unit specifically for, described noise to be detected is carried out pre-emphasis, framing and adds Hamming window process; For each frame signal of described noise to be detected, carry out Fourier transform respectively, obtain the amplitude spectrum of each frame signal described; The second feature that logarithm operation and discrete cosine transform obtain described noise to be detected is done to described amplitude spectrum.
In conjunction with the embodiment that the second of second aspect is possible, embodiments provide the third possible embodiment of second aspect, wherein, described acquisition unit also for, in each frame signal for described noise to be detected, carry out Fourier transform respectively, after obtaining the amplitude spectrum of each frame signal described, filtering is carried out to described amplitude spectrum.
In conjunction with second aspect, or the first of second aspect, the second or the third possible embodiment, embodiments provide the 4th kind of possible embodiment of second aspect, wherein, described choose unit specifically for, Hidden Markov Model (HMM) is adopted the sample information in described information to be matched and described standard feature Sample Storehouse to be contrasted, from described standard feature Sample Storehouse, obtain the sample information reaching predetermined threshold value with described information matches degree to be matched, using the sound source type of the sound source type of the noise of the known sound source type corresponding with the sample information that described information matches degree to be matched reaches predetermined threshold value as described noise to be detected.
The method and apparatus provided in the embodiment of the present invention, abandon in prior art the pattern adopting BP (Back Propagation) neural network to carry out identification of sound source, the sample information selecting vector quantization to obtain dexterously carries out identification of sound source, greatly reduce computation complexity, significantly improve noise source recognition efficiency.
Further, the Mel frequency cepstral coefficient MFCC feature preferably extracted in noise in the embodiment of the present invention is basic as subsequent treatment, MFCC feature can reflect the aural signature of people's ear preferably, makes final recognition result more meet the actual impression of people, realistic demand.
The method that the embodiment of the present invention provides and device, it is convenient to implement, and has outstanding substantive distinguishing features and marked improvement, is applicable to large-scale promotion application.
For making above-mentioned purpose of the present invention, feature and advantage become apparent, preferred embodiment cited below particularly, and coordinate appended accompanying drawing, be described in detail below.
Accompanying drawing explanation
In order to be illustrated more clearly in the technical scheme of the embodiment of the present invention, be briefly described to the accompanying drawing used required in embodiment below, be to be understood that, the following drawings illustrate only some embodiment of the present invention, therefore the restriction to scope should be counted as, for those of ordinary skill in the art, under the prerequisite not paying creative work, other relevant accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1 shows a kind of schematic flow sheet that the embodiment of the present invention 1 provides;
Fig. 2 shows a kind of schematic flow sheet obtaining second feature that the embodiment of the present invention 1 provides;
The another kind that Fig. 3 shows the embodiment of the present invention 1 to be provided obtains the schematic flow sheet of second feature;
Fig. 4 shows the another kind of structured flowchart that the embodiment of the present invention 2 provides;
Fig. 5 shows the another kind of structured flowchart that the embodiment of the present invention 2 provides.
Embodiment
Below in conjunction with accompanying drawing in the embodiment of the present invention, be clearly and completely described the technical scheme in the embodiment of the present invention, obviously, described embodiment is only the present invention's part embodiment, instead of whole embodiments.The assembly of the embodiment of the present invention describing and illustrate in usual accompanying drawing herein can be arranged with various different configuration and design.Therefore, below to the detailed description of the embodiments of the invention provided in the accompanying drawings and the claimed scope of the present invention of not intended to be limiting, but selected embodiment of the present invention is only represented.Based on embodiments of the invention, the every other embodiment that those skilled in the art obtain under the prerequisite not making creative work, all belongs to the scope of protection of the invention.
Embodiment 1
Now, mainly based on the BP trained (Back Propagation) neural network, BP neural network corresponding for noise source to be identified and the BP neural network trained are contrasted, identify noise source, inventor finds after deliberation, this recognition method computation complexity is comparatively large, has had a strong impact on noise source recognition efficiency.Based on this, as shown in Figure 1, embodiments provide a kind of neighbourhood noise identification of sound source method, be applied to neighbourhood noise identification of sound source device, standard feature Sample Storehouse is preset with in described neighbourhood noise identification of sound source device, described standard feature Sample Storehouse comprises and the fisrt feature of the noise of known sound source type is carried out by preset rules the sample information that vector quantization obtains, and described sample information is corresponding with the noise of described known sound source type, and described method comprises:
Step S100: described neighbourhood noise identification of sound source device obtains the second feature of noise to be detected;
In order to reflect the actual impression of people's ear to noise better, wherein, fisrt feature and second feature all preferably Mel frequency cepstral coefficient MFCC feature.Wherein, the mode obtaining MFCC feature has multiple, provides wherein a kind of MFCC feature obtain manner in the embodiment of the present invention, as shown in Figure 2, and step S200: described noise to be detected is carried out pre-emphasis, framing and adds Hamming window process;
Wherein, pre-emphasis is that noise signal is done high-pass filtering process, and the design of wave filter has multiple, preferably adopts following design in the present embodiment:
H (z)=1-μ z
-1, wherein, the value of μ, between 0.9 ~ 1.0, preferably gets 0.97.
Framing is processing procedure common in Speech processing, and in the present embodiment, the sampling number N comprised in every frame is relevant with sample frequency with every frame coincidence sampling number M, and preferably the every frame time length of guarantee is 20ms ~ 30ms, M value is 1/3 to 1/2 of N.Such as: when sample frequency is 16KHz, N gets 512, M and gets 192.Follow-up processing procedure is all carry out separately for every frame sampling point.
The form adding Hamming window has multiple, preferably adopts the Hamming window of following form in the present embodiment:
Wherein, N represents a frame data sampled point number, and pi is circular constant.
Step S201: for each frame signal of described noise to be detected, carry out Fourier transform respectively, obtains the amplitude spectrum of each frame signal described;
Wherein, preferably Fast Fourier Transform (FFT) FFT is adopted.
Step S203: the second feature that logarithm operation and discrete cosine transform obtain described noise to be detected is done to described amplitude spectrum.
Wherein, the form of discrete cosine transform has multiple, preferably adopts following form in the present embodiment:
Wherein, C (n) i.e. discrete cosine transform, s (m) represents wave filter and exports, and N represents a frame data sampled point number, and M refers to Mel number of filter, and L is MFCC coefficient exponent number, preferably gets 13.
In order to ensure the accuracy that second feature obtains, as shown in Figure 3, preferably in each frame signal for described noise to be detected, carry out Fourier transform respectively, after obtaining the amplitude spectrum of each frame signal described, perform step S202: filtering is carried out to described amplitude spectrum, in the present embodiment, preferably carries out Mel filtering; Then step S203 is performed: the second feature that logarithm operation and discrete cosine transform obtain described noise to be detected is done to filtered described amplitude spectrum.
The obtain manner of the fisrt feature of the noise of known sound source type is identical with the obtain manner of second feature, does not do repeat specification at this.
Step S101: by described preset rules, the described second feature obtained is carried out vector quantization and obtain information to be matched;
Wherein, by the preset rules obtaining sample information, the second feature of noise to be detected is carried out vector quantization and obtain information to be matched, namely the noise of known sound source type and the noise to be detected of unknown sound source type are adopted and are obtained sample information and information to be matched respectively in a like fashion, it should be noted that, standard feature Sample Storehouse comprises the sample information of the multiple noise substantially containing all neighbourhood noise sound source types.
Step S102: the sample information in described information to be matched and described standard feature Sample Storehouse is contrasted, from described standard feature Sample Storehouse, obtain the sample information reaching predetermined threshold value with described information matches degree to be matched, using the sound source type of the sound source type of the noise of the known sound source type corresponding with the sample information that described information matches degree to be matched reaches predetermined threshold value as described noise to be detected.
Wherein, the sample information in information to be matched and standard feature Sample Storehouse contrasted, the mode obtaining the sound source type of noise to be detected has multiple, such as: adopt Hidden Markov Model (HMM) to obtain.Again such as: set after carrying out vector quantization for (multiple) fisrt feature, constitute hyperspace, the noise characteristic of different sound source type has different distributions in hyperspace, namely there is different Codebook of Vector Quantization, Codebook of Vector Quantization comprises the reference point that several the bests represent this distribution, the Codebook of Vector Quantization of every noise like the second feature of noise to be detected is utilized to carry out vector quantization respectively, each vector quantization all calculates corresponding quantization error, quantization error is lower than (maximum error) predetermined threshold value, namely when matching degree reaches (minimum matching degree) predetermined threshold value, by the sound source type of corresponding sound source type as noise to be detected.According to the actual requirements, also can quantization error is minimum time corresponding noise type as the sound source type of noise to be detected.
The embodiment of the present invention has been abandoned in prior art and has been adopted BP neural network to carry out the pattern of identification of sound source, and the sample information selecting vector quantization to obtain dexterously carries out identification of sound source, greatly reduces computation complexity, significantly improves noise source recognition efficiency; Mel frequency cepstral coefficient MFCC feature in preferred extraction noise is as subsequent treatment basis, and MFCC feature can reflect the aural signature of people's ear preferably, makes final recognition result more meet the actual impression of people, realistic demand.
Embodiment 2
Now, mainly based on the BP trained (Back Propagation) neural network, BP neural network corresponding for noise source to be identified and the BP neural network trained are contrasted, identify noise source, inventor finds after deliberation, this recognition method computation complexity is comparatively large, has had a strong impact on noise source recognition efficiency.
Based on this, as shown in Figure 4, embodiments provide a kind of neighbourhood noise identification of sound source device, comprise: preset unit, for preset standard feature samples storehouse, described standard feature Sample Storehouse comprises and the fisrt feature of the noise of known sound source type is carried out by preset rules the sample information that vector quantization obtains, and described sample information is corresponding with the noise of described known sound source type; Obtain unit, for obtaining the second feature of noise to be detected; Vector quantization unit, obtains information to be matched for the described second feature that described acquisition unit obtains being carried out vector quantization by described preset rules; Choose unit, for the sample information in described information to be matched and described standard feature Sample Storehouse is contrasted, from described standard feature Sample Storehouse, obtain the sample information reaching predetermined threshold value with described information matches degree to be matched, using the sound source type of the sound source type of the noise of the known sound source type corresponding with the sample information that described information matches degree to be matched reaches predetermined threshold value as described noise to be detected.
In order to reflect the actual impression of people's ear to noise better, wherein, fisrt feature and second feature all preferably Mel frequency cepstral coefficient MFCC feature.Wherein, the mode obtaining MFCC feature has multiple, provides wherein a kind ofly to obtain the mode that unit obtains MFCC feature in the embodiment of the present invention, described acquisition unit specifically for, described noise to be detected is carried out pre-emphasis, framing and adds Hamming window process; For each frame signal of described noise to be detected, carry out Fourier transform respectively, obtain the amplitude spectrum of each frame signal described; The second feature that logarithm operation and discrete cosine transform obtain described noise to be detected is done to described amplitude spectrum.
In order to ensure the accuracy that second feature obtains, obtain unit also for, in each frame signal for described noise to be detected, carry out Fourier transform respectively, after obtaining the amplitude spectrum of each frame signal described, filtering carried out to described amplitude spectrum.
Wherein, choose unit and the sample information in information to be matched and standard feature Sample Storehouse contrasted, the mode obtaining the sound source type of noise to be detected has multiple, such as: adopt Hidden Markov Model (HMM) to obtain.Again such as: set after carrying out vector quantization for (multiple) fisrt feature, constitute hyperspace, the noise characteristic of different sound source type has different distributions in hyperspace, namely there is different Codebook of Vector Quantization, Codebook of Vector Quantization comprises the reference point that several the bests represent this distribution, the Codebook of Vector Quantization of every noise like the second feature of noise to be detected is utilized to carry out vector quantization respectively, each vector quantization all calculates corresponding quantization error, quantization error is lower than (maximum error) predetermined threshold value, namely when matching degree reaches (minimum matching degree) predetermined threshold value, by the sound source type of corresponding sound source type as noise to be detected.According to the actual requirements, also can quantization error is minimum time corresponding noise type as the sound source type of noise to be detected.
The embodiment of the present invention has been abandoned in prior art and has been adopted BP neural network to carry out the pattern of identification of sound source, and the sample information selecting vector quantization to obtain dexterously carries out identification of sound source, greatly reduces computation complexity, significantly improves noise source recognition efficiency; Mel frequency cepstral coefficient MFCC feature in preferred extraction noise is as subsequent treatment basis, and MFCC feature can reflect the aural signature of people's ear preferably, makes final recognition result more meet the actual impression of people, realistic demand.
The device that the embodiment of the present invention provides, its technique effect realizing principle and generation is identical with preceding method embodiment, is concise and to the point description, and the not mentioned part of device embodiment part can with reference to corresponding contents in preceding method embodiment.
As shown in Figure 5, the embodiment of the present invention additionally provides a kind of structural representation of neighbourhood noise identification of sound source device, comprising: processor 400, storer 404, bus 402 and communication interface 403, described processor 400, communication interface 403 are connected by bus 402 with storer 404; .
Wherein, storer 404 may comprise high-speed random access memory (RAM:Random Access Memory), also non-labile storer (non-volatile memory) may also be comprised, such as at least one magnetic disk memory.Realize the communication connection between this system network element and at least one other network element by least one communication interface 403 (can be wired or wireless), can internet be used, wide area network, local network, Metropolitan Area Network (MAN) etc.
Processor 400 for the executable module in execute store 404, such as computer program 401; Processor 400 is by communication interface 403 receiving data stream;
Bus 402 can be isa bus, pci bus or eisa bus etc.Described bus can be divided into address bus, data bus, control bus etc.For ease of representing, only representing with a four-headed arrow in Fig. 5, but not representing the bus only having a bus or a type.
Wherein, storer 404 is for storage program 401, and described processor 400, after receiving execution instruction, performs described program 401, the method performed by device of the procedure definition that aforementioned embodiment of the present invention any embodiment discloses can be applied in processor 400, or is realized by processor 400.
In specific implementation, program 401 can comprise program code, and described program code comprises computer-managed instruction and algorithm etc.;
Processor 400 may be a kind of integrated circuit (IC) chip, has the processing power of signal.In implementation procedure, each step of said method can be completed by the instruction of the integrated logic circuit of the hardware in processor 400 or software form.Above-mentioned processor 400 can be general processor, comprises central processing unit (Central Processing Unit is called for short CPU), network processing unit (Network Processor is called for short NP) etc.; Can also be digital signal processor (DSP), special IC (ASIC), ready-made programmable gate array (FPGA) or other programmable logic device (PLD), discrete gate or transistor logic, discrete hardware components.Can realize or perform disclosed each method, step and the logic diagram in the embodiment of the present invention.The processor etc. of general processor can be microprocessor or this processor also can be any routine.Step in conjunction with the method disclosed in the embodiment of the present invention directly can be presented as that hardware decoding processor is complete, or combines complete by the hardware in decoding processor and software module.Software module can be positioned at random access memory, flash memory, ROM (read-only memory), in the storage medium of this area maturations such as programmable read only memory or electrically erasable programmable storer, register.This storage medium is positioned at storer 404, and processor 400 reads the information in storer 404, completes the step of said method in conjunction with its hardware.
The computer program that what the embodiment of the present invention provided carry out in device, comprise the computer-readable recording medium storing program code, the instruction that described program code comprises can be used for performing the method described in previous methods embodiment, specific implementation see embodiment of the method, can not repeat them here.
Those skilled in the art can be well understood to, and for convenience and simplicity of description, the device of foregoing description and the specific works process of unit, with reference to the corresponding process in preceding method embodiment, can not repeat them here.
In several embodiments that the application provides, should be understood that disclosed apparatus and method can realize by another way.Device embodiment described above is only schematic, and such as, the process flow diagram in accompanying drawing and block diagram show device according to multiple embodiment of the present invention, the architectural framework in the cards of method and computer program product, function and operation.In this, each square frame in process flow diagram or block diagram can represent a part for module, program segment or a code, and a part for described module, program segment or code comprises one or more executable instruction for realizing the logic function specified.Also it should be noted that at some as in the realization of replacing, the function marked in square frame also can be different from occurring in sequence of marking in accompanying drawing.Such as, in fact two continuous print square frames can perform substantially concurrently, and they also can perform by contrary order sometimes, and this determines according to involved function.Also it should be noted that, the combination of the square frame in each square frame in block diagram and/or process flow diagram and block diagram and/or process flow diagram, can realize by the special hardware based system of the function put rules into practice or action, or can realize with the combination of specialized hardware and computer instruction.
The described unit illustrated as separating component or can may not be and physically separates, and the parts as unit display can be or may not be physical location, namely can be positioned at a place, or also can be distributed in multiple network element.Some or all of unit wherein can be selected according to the actual needs to realize the object of the present embodiment scheme.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing unit, also can be that the independent physics of unit exists, also can two or more unit in a unit integrated.
If described function using the form of SFU software functional unit realize and as independently production marketing or use time, can be stored in a computer read/write memory medium.Based on such understanding, the part of the part that technical scheme of the present invention contributes to prior art in essence in other words or this technical scheme can embody with the form of software product, this computer software product is stored in a storage medium, comprising some instructions in order to make a computer equipment (can be personal computer, server, or the network equipment etc.) perform all or part of step of method described in each embodiment of the present invention.And aforesaid storage medium comprises: USB flash disk, portable hard drive, ROM (read-only memory) (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disc or CD etc. various can be program code stored medium.
The above; be only the specific embodiment of the present invention, but protection scope of the present invention is not limited thereto, is anyly familiar with those skilled in the art in the technical scope that the present invention discloses; change can be expected easily or replace, all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should described be as the criterion with the protection domain of claim.
Claims (10)
1. a neighbourhood noise identification of sound source method, it is characterized in that, be applied to neighbourhood noise identification of sound source device, standard feature Sample Storehouse is preset with in described neighbourhood noise identification of sound source device, described standard feature Sample Storehouse comprises and the fisrt feature of the noise of known sound source type is carried out by preset rules the sample information that vector quantization obtains, described sample information is corresponding with the noise of described known sound source type, and described method comprises:
Described neighbourhood noise identification of sound source device obtains the second feature of noise to be detected;
By described preset rules, the described second feature obtained is carried out vector quantization and obtain information to be matched;
Sample information in described information to be matched and described standard feature Sample Storehouse is contrasted, from described standard feature Sample Storehouse, obtain the sample information reaching predetermined threshold value with described information matches degree to be matched, using the sound source type of the sound source type of the noise of the known sound source type corresponding with the sample information that described information matches degree to be matched reaches predetermined threshold value as described noise to be detected.
2. neighbourhood noise identification of sound source method according to claim 1, is characterized in that, described fisrt feature and second feature are Mel frequency cepstral coefficient MFCC feature.
3. neighbourhood noise identification of sound source method according to claim 1, is characterized in that, described neighbourhood noise identification of sound source device obtains the second feature of noise to be detected, comprising:
Described noise to be detected is carried out pre-emphasis, framing and adds Hamming window process;
For each frame signal of described noise to be detected, carry out Fourier transform respectively, obtain the amplitude spectrum of each frame signal described;
The second feature that logarithm operation and discrete cosine transform obtain described noise to be detected is done to described amplitude spectrum.
4. neighbourhood noise identification of sound source method according to claim 3, is characterized in that, described each frame signal for described noise to be detected, carries out Fourier transform respectively, after obtaining the amplitude spectrum of each frame signal described, also comprises:
Filtering is carried out to described amplitude spectrum;
Described the second feature that logarithm operation and discrete cosine transform obtain described noise to be detected is done to described amplitude spectrum, comprising:
The second feature that logarithm operation and discrete cosine transform obtain described noise to be detected is done to filtered described amplitude spectrum.
5. the neighbourhood noise identification of sound source method according to Claims 1 to 4 any one, it is characterized in that, described sample information in described information to be matched and described standard feature Sample Storehouse to be contrasted, from described standard feature Sample Storehouse, obtain the sample information reaching predetermined threshold value with described information matches degree to be matched, comprising:
Adopt Hidden Markov Model (HMM) the sample information in described information to be matched and described standard feature Sample Storehouse to be contrasted, from described standard feature Sample Storehouse, obtain the sample information reaching predetermined threshold value with described information matches degree to be matched.
6. a neighbourhood noise identification of sound source device, is characterized in that, comprising:
Preset unit, for preset standard feature samples storehouse, described standard feature Sample Storehouse comprises and the fisrt feature of the noise of known sound source type is carried out by preset rules the sample information that vector quantization obtains, and described sample information is corresponding with the noise of described known sound source type;
Obtain unit, for obtaining the second feature of noise to be detected;
Vector quantization unit, obtains information to be matched for the described second feature that described acquisition unit obtains being carried out vector quantization by described preset rules;
Choose unit, for the sample information in described information to be matched and described standard feature Sample Storehouse is contrasted, from described standard feature Sample Storehouse, obtain the sample information reaching predetermined threshold value with described information matches degree to be matched, using the sound source type of the sound source type of the noise of the known sound source type corresponding with the sample information that described information matches degree to be matched reaches predetermined threshold value as described noise to be detected.
7. neighbourhood noise identification of sound source device according to claim 6, is characterized in that, described fisrt feature and second feature are Mel frequency cepstral coefficient MFCC feature.
8. neighbourhood noise identification of sound source device according to claim 6, is characterized in that, described acquisition unit specifically for, described noise to be detected is carried out pre-emphasis, framing and adds Hamming window process; For each frame signal of described noise to be detected, carry out Fourier transform respectively, obtain the amplitude spectrum of each frame signal described; The second feature that logarithm operation and discrete cosine transform obtain described noise to be detected is done to described amplitude spectrum.
9. neighbourhood noise identification of sound source device according to claim 8, is characterized in that, described acquisition unit also for, in each frame signal for described noise to be detected, carry out Fourier transform respectively, after obtaining the amplitude spectrum of each frame signal described, filtering is carried out to described amplitude spectrum.
10. the neighbourhood noise identification of sound source device according to claim 6 ~ 9 any one, it is characterized in that, described choose unit specifically for, Hidden Markov Model (HMM) is adopted the sample information in described information to be matched and described standard feature Sample Storehouse to be contrasted, from described standard feature Sample Storehouse, obtain the sample information reaching predetermined threshold value with described information matches degree to be matched, using the sound source type of the sound source type of the noise of the known sound source type corresponding with the sample information that described information matches degree to be matched reaches predetermined threshold value as described noise to be detected.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510304135.1A CN104952449A (en) | 2015-01-09 | 2015-06-04 | Method and device for identifying environmental noise sources |
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510012149 | 2015-01-09 | ||
| CN2015100121496 | 2015-01-09 | ||
| CN201510304135.1A CN104952449A (en) | 2015-01-09 | 2015-06-04 | Method and device for identifying environmental noise sources |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104952449A true CN104952449A (en) | 2015-09-30 |
Family
ID=54167061
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510304135.1A Pending CN104952449A (en) | 2015-01-09 | 2015-06-04 | Method and device for identifying environmental noise sources |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104952449A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105632513A (en) * | 2015-12-18 | 2016-06-01 | 合肥寰景信息技术有限公司 | Network community voice filtering method |
| CN105678256A (en) * | 2016-01-06 | 2016-06-15 | 南方科技大学 | Signal processing method, signal processing device and signal processing system |
| CN106251868A (en) * | 2016-08-09 | 2016-12-21 | 江门雷斯诺照明有限公司 | A kind of light fixture voice identification control method with Intelligent noise reduction function |
| CN107393523A (en) * | 2017-07-28 | 2017-11-24 | 深圳市盛路物联通讯技术有限公司 | A kind of noise monitoring method and system |
| CN107967917A (en) * | 2016-10-19 | 2018-04-27 | 福特全球技术公司 | The vehicle periphery audio classification learnt by neural network machine |
| CN108257617A (en) * | 2018-01-11 | 2018-07-06 | 会听声学科技（北京）有限公司 | A kind of noise scenarios identifying system and method |
| WO2018161429A1 (en) * | 2017-03-07 | 2018-09-13 | 华为技术有限公司 | Noise detection method, and terminal apparatus |
| CN111832604A (en) * | 2020-05-21 | 2020-10-27 | 中煤科工集团重庆研究院有限公司 | Acoustic emission signal noise filtering and danger comprehensive judgment method based on multi-system fusion |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060253282A1 (en) * | 2005-03-14 | 2006-11-09 | Schmidt Gerhard U | System for automatic recognition of vehicle operating noises |
| CN101051462A (en) * | 2006-04-07 | 2007-10-10 | 株式会社东芝 | Feature-vector compensating apparatus and feature-vector compensating method |
| CN101980336A (en) * | 2010-10-18 | 2011-02-23 | 福州星网视易信息系统有限公司 | Hidden Markov model-based vehicle sound identification method |
| CN102231277A (en) * | 2011-06-29 | 2011-11-02 | 电子科技大学 | Method for protecting mobile terminal privacy based on voiceprint recognition |
| CN102539154A (en) * | 2011-10-16 | 2012-07-04 | 浙江吉利汽车研究院有限公司 | Engine fault diagnosis method and device based on exhaust noise vector quantitative analysis |
| CN102739863A (en) * | 2012-06-14 | 2012-10-17 | 中兴通讯股份有限公司 | Emergency call method and device |
| CN103426439A (en) * | 2013-05-08 | 2013-12-04 | 杭州联汇数字科技有限公司 | Method for detecting consistency of broadcast television audio signal content |
| CN103456301A (en) * | 2012-05-28 | 2013-12-18 | 中兴通讯股份有限公司 | Ambient sound based scene recognition method and device and mobile terminal |
| CN103971680A (en) * | 2013-01-24 | 2014-08-06 | 华为终端有限公司 | Method and device for recognizing voices |
| CN104167207A (en) * | 2014-06-20 | 2014-11-26 | 国家电网公司 | Equipment sound identification method based on transformer substation patrol inspection robot |
-
2015
- 2015-06-04 CN CN201510304135.1A patent/CN104952449A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060253282A1 (en) * | 2005-03-14 | 2006-11-09 | Schmidt Gerhard U | System for automatic recognition of vehicle operating noises |
| CN101051462A (en) * | 2006-04-07 | 2007-10-10 | 株式会社东芝 | Feature-vector compensating apparatus and feature-vector compensating method |
| CN101980336A (en) * | 2010-10-18 | 2011-02-23 | 福州星网视易信息系统有限公司 | Hidden Markov model-based vehicle sound identification method |
| CN102231277A (en) * | 2011-06-29 | 2011-11-02 | 电子科技大学 | Method for protecting mobile terminal privacy based on voiceprint recognition |
| CN102539154A (en) * | 2011-10-16 | 2012-07-04 | 浙江吉利汽车研究院有限公司 | Engine fault diagnosis method and device based on exhaust noise vector quantitative analysis |
| CN103456301A (en) * | 2012-05-28 | 2013-12-18 | 中兴通讯股份有限公司 | Ambient sound based scene recognition method and device and mobile terminal |
| CN102739863A (en) * | 2012-06-14 | 2012-10-17 | 中兴通讯股份有限公司 | Emergency call method and device |
| CN103971680A (en) * | 2013-01-24 | 2014-08-06 | 华为终端有限公司 | Method and device for recognizing voices |
| CN103426439A (en) * | 2013-05-08 | 2013-12-04 | 杭州联汇数字科技有限公司 | Method for detecting consistency of broadcast television audio signal content |
| CN104167207A (en) * | 2014-06-20 | 2014-11-26 | 国家电网公司 | Equipment sound identification method based on transformer substation patrol inspection robot |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105632513A (en) * | 2015-12-18 | 2016-06-01 | 合肥寰景信息技术有限公司 | Network community voice filtering method |
| CN105678256B (en) * | 2016-01-06 | 2019-01-18 | 南方科技大学 | Signal processing method, signal processing apparatus and signal processing system |
| WO2017118189A1 (en) * | 2016-01-06 | 2017-07-13 | 南方科技大学 | Signal processing method, signal processing apparatus and signal processing system |
| CN105678256A (en) * | 2016-01-06 | 2016-06-15 | 南方科技大学 | Signal processing method, signal processing device and signal processing system |
| CN106251868A (en) * | 2016-08-09 | 2016-12-21 | 江门雷斯诺照明有限公司 | A kind of light fixture voice identification control method with Intelligent noise reduction function |
| CN107967917A (en) * | 2016-10-19 | 2018-04-27 | 福特全球技术公司 | The vehicle periphery audio classification learnt by neural network machine |
| WO2018161429A1 (en) * | 2017-03-07 | 2018-09-13 | 华为技术有限公司 | Noise detection method, and terminal apparatus |
| CN109074814A (en) * | 2017-03-07 | 2018-12-21 | 华为技术有限公司 | A kind of noise detecting method and terminal device |
| CN109074814B (en) * | 2017-03-07 | 2023-05-09 | 华为技术有限公司 | Noise detection method and terminal equipment |
| CN107393523A (en) * | 2017-07-28 | 2017-11-24 | 深圳市盛路物联通讯技术有限公司 | A kind of noise monitoring method and system |
| CN107393523B (en) * | 2017-07-28 | 2020-11-13 | 深圳市盛路物联通讯技术有限公司 | Noise monitoring method and system |
| CN108257617A (en) * | 2018-01-11 | 2018-07-06 | 会听声学科技（北京）有限公司 | A kind of noise scenarios identifying system and method |
| CN111832604A (en) * | 2020-05-21 | 2020-10-27 | 中煤科工集团重庆研究院有限公司 | Acoustic emission signal noise filtering and danger comprehensive judgment method based on multi-system fusion |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104952449A (en) | Method and device for identifying environmental noise sources | |
| Lee et al. | Sample-level deep convolutional neural networks for music auto-tagging using raw waveforms | |
| CN110120224B (en) | Method and device for constructing bird sound recognition model, computer equipment and storage medium | |
| CN107610707B (en) | A kind of method for recognizing sound-groove and device | |
| US8831942B1 (en) | System and method for pitch based gender identification with suspicious speaker detection | |
| US20190172442A1 (en) | Method for forming the excitation signal for a glottal pulse model based parametric speech synthesis system | |
| CN111048071B (en) | Voice data processing method, device, computer equipment and storage medium | |
| WO2012134541A1 (en) | Speech syllable/vowel/phone boundary detection using auditory attention cues | |
| CN101136199A (en) | Voice data processing method and equipment | |
| CN101894548A (en) | Modeling method and modeling device for language identification | |
| CN106294331A (en) | Audio information retrieval method and device | |
| CN109767756A (en) | A kind of speech feature extraction algorithm based on dynamic partition inverse discrete cosine transform cepstrum coefficient | |
| CN102982351A (en) | Porcelain insulator vibrational acoustics test data sorting technique based on back propagation (BP) neural network | |
| Jung et al. | DNN based multi-level feature ensemble for acoustic scene classification. | |
| CN112599148A (en) | Voice recognition method and device | |
| CA3004700A1 (en) | Method for forming the excitation signal for a glottal pulse model based parametric speech synthesis system | |
| Liu et al. | Simple pooling front-ends for efficient audio classification | |
| CN106297768B (en) | Speech recognition method | |
| CN113921017A (en) | Voice identity detection method and device, electronic equipment and storage medium | |
| Nilsson et al. | On the mutual information between frequency bands in speech | |
| CN110226201A (en) | The voice recognition indicated using the period | |
| Nirjon et al. | sMFCC: exploiting sparseness in speech for fast acoustic feature extraction on mobile devices--a feasibility study | |
| Sundar et al. | A mixture model approach for formant tracking and the robustness of student's-t distribution | |
| Prasanna Kumar et al. | Single-channel speech separation using combined EMD and speech-specific information | |
| Mini et al. | Feature vector selection of fusion of MFCC and SMRT coefficients for SVM classifier based speech recognition system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20150914 Address after: Pingtung 519000 Guangdong province Zhuhai Nanping Science and Technology Industrial Park Road No. Applicant after: Zhuhai Comleader Information Science & Technology Co., Ltd. Address before: Pingtung 519000 Guangdong Province, Zhuhai city Xiangzhou District Nanping Technology Industrial Park Road No. 1 A225 Applicant before: ZHUHAI GAOLING TECHNOLOGY CO., LTD. |
|
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150930 |
|
| RJ01 | Rejection of invention patent application after publication |