CN105812068B - A kind of noise suppressing method and device based on Gaussian Profile weighting - Google Patents
A kind of noise suppressing method and device based on Gaussian Profile weighting Download PDFInfo
- Publication number
- CN105812068B CN105812068B CN201610169563.2A CN201610169563A CN105812068B CN 105812068 B CN105812068 B CN 105812068B CN 201610169563 A CN201610169563 A CN 201610169563A CN 105812068 B CN105812068 B CN 105812068B
- Authority
- CN
- China
- Prior art keywords
- signal
- radius
- wave filter
- window
- noise
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000012545 processing Methods 0.000 claims abstract description 17
- 238000001914 filtration Methods 0.000 claims description 15
- 230000001629 suppression Effects 0.000 claims description 11
- 230000008030 elimination Effects 0.000 claims description 3
- 238000003379 elimination reaction Methods 0.000 claims description 3
- 230000011664 signaling Effects 0.000 claims description 3
- 230000006870 function Effects 0.000 description 44
- 238000009826 distribution Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000012544 monitoring process Methods 0.000 description 5
- 239000002245 particle Substances 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012952 Resampling Methods 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000004615 ingredient Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B15/00—Suppression or limitation of noise or interference
- H04B15/005—Reducing noise, e.g. humm, from the supply
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
- G10L19/04—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using predictive techniques
- G10L19/26—Pre-filtering or post-filtering
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0208—Noise filtering
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Human Computer Interaction (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Computer Networks & Wireless Communication (AREA)
- Quality & Reliability (AREA)
- Image Processing (AREA)
- Noise Elimination (AREA)
Abstract
The present invention relates to a kind of noise suppressing method and device based on Gaussian Profile weighting, wherein, noise suppressing method includes：Determine the local radius of wave filter；The length of window of wave filter, and discrete signal in the window of definite wave filter are determined according to the local radius of the wave filter；Corresponding average and variance are determined according to discrete signal in the window of the wave filter；The Gaussian function of discrete signal in the window of wave filter is determined using the average and variance；The corresponding Gaussian function numerical value of each discrete signal in the window of wave filter is determined using the Gaussian function, and is summed to the Gaussian function numerical value of acquisition；Utilize the corresponding Gaussian function numerical value of each discrete signal in the window of wave filter and the weighted value corresponding with each discrete signal in the window of definite wave filter of Gaussian function numerical value；The wave filter carries out noise suppressed processing using weighted value to the discrete signal of window center.
Description
Technical field
The present invention relates to noise management technique field, more particularly to a kind of noise suppressing method based on Gaussian Profile weighting
And device.
Background technology
Signal filtering technique is the core research topic of field of signal processing.Signal filtering method is divided into linear filtering and non-
Linear filtering.In early stage digital signal and picture signal processing and studying, linear filter technology is the primary hand for suppressing noise
Section, this has appropriate mathematical expression form mainly due to linear filtering mode and easily designs and Implements.When linear filtering skill
When art is applied to nonadditivity noise signal, its result is often not satisfied.Signal collection, in transmitting procedure unavoidably by
To different degrees of noise jamming, impact signal is even produced sometimes.At this time, it may be necessary to appropriate processing is carried out to signal, to disappear
Except impact and its noise contribution.When being filtered using linear filter technology, its effect is often barely satisfactory, can not obtain
Obtain preferable effect.
Medium filtering is a kind of nonlinear signal processing technology that can effectively suppress noise based on sequencing statistical theory.It is right
For signal, it is ranked up the data in neighborhood, using the median in neighborhood as currency.This method can have
Effect suppresses impact noise, but causes signal not smooth enough after filtering.
Mean filter is to be filtered linear filter to signal based on signal local statistic information, equivalent to one low pass
Wave filter.The establishment of this method algorithm is convenient, and execution speed is fast, and this method is while realizing that signal smoothing suppresses noise, easily
The details of blurred signal.
Particle filter is to carry out approximation to probability density function by finding one group of random sample propagated in state space,
Integral operation is replaced with sample average, so as to obtain the process of minimum variance distribution.Particle filter can be expressed more accurately
Posterior probability distribution based on observed quantity and controlled quentity controlled variable.But its main problem is to need substantial amounts of sample size could be near well
Like the posterior probability function of system.Resampling technique is used additionally, due to algorithm, this can cause sample availability and multifarious
Loss, causes samples impoverishment phenomenon.
Common other wave filters further include the nonlinear filters such as exponent filtering, shape filtering, weighted filtering.These filters
Ripple algorithm is suitable for different aspects, respectively there is its quality.For in signal acquisition process in Hydropower Unit status monitoring field
The noise jamming of generation, even impulsive disturbance, all the time there has been no good solution, cause Hydropower Unit runout to protect
Protecting system such as fails to come into operation completely at the series of problems.
The content of the invention
The main purpose of the embodiment of the present invention is to propose a kind of noise suppressing method and dress based on Gaussian Profile weighting
Put, realize the suppression that ingredient noise is even impacted containing Gaussian noise, make obtained signal truer.
To achieve the above object, the present invention provides a kind of noise suppressing method based on Gaussian Profile weighting, including：
Determine the local radius of wave filter；
The length of window of wave filter is determined according to the local radius of the wave filter, and it is discrete in the window of definite wave filter
Signal；
The corresponding average of current local radius and variance are determined according to discrete signal in the window of the wave filter；
The Gaussian function of discrete signal in the window of wave filter is determined using the average and variance；
The corresponding Gaussian function numerical value of each discrete signal in the window of wave filter is determined using the Gaussian function, and to obtaining
The Gaussian function numerical value summation obtained；
Using the corresponding Gaussian function numerical value of each discrete signal in the window of wave filter and Gaussian function numerical value and determine
The corresponding weighted value of each discrete signal in the window of wave filter；
The wave filter carries out noise suppressed processing using weighted value to the discrete signal of window center.
Preferably, the step of local radius of the definite wave filter includes：
When data index value i is less than optimal local radius, then the local radius of wave filter is equal to data index value i；
When data index value i is more than or equal to optimal local radius, and data index value i is less than the number of discrete signal and subtracts
During optimal local radius, then, the local radius of wave filter is equal to optimal local radius；
Otherwise, the local radius of wave filter is equal to the number of discrete signal and subtracts current data index value and subtract 1 again.
Preferably, the local radius of the wave filter of the length of window of the wave filter equal to twice adds 1 again.
Preferably, the step of wave filter carries out noise suppressed processing using weighted value to the discrete signal of window center
Including：
Discrete signal in window is carried out dot-product operation by the wave filter with corresponding weighted value, which is filter
Ripple device exports the noise suppressed result of currency.
Preferably, the optimal local radius obtaining step includes：
Noise is added to emulation signal；
Noise suppressed processing is carried out to the emulation signal after addition noise, obtains noise cancellation signal；
Obtain the mean square error between emulation signal and the noise cancellation signal corresponding to current radius；
To corresponding to the mean square error between the emulation signal corresponding to current radius and noise cancellation signal and upper Radius
Signal is emulated compared with the mean square error between noise cancellation signal；If the emulation signal corresponding to current radius is believed with de-noising
Mean square error between number is more than or equal to the mean square error between emulation signal and noise cancellation signal corresponding to upper Radius, then when
Preceding radius is optimal local radius；Otherwise, lower Radius is relatively more current as upper Radius as current radius, current radius
The mean square error emulated between signal and noise cancellation signal and the emulation signal corresponding to upper Radius and de-noising corresponding to radius
Mean square error between signal, until obtaining optimal local radius.
Accordingly, to achieve the above object, present invention also offers a kind of noise suppressed dress based on Gaussian Profile weighting
Put, including：
Local radius determination unit, for determining the local radius of wave filter；
Discrete signal determination unit in window, the window for determining wave filter according to the local radius of the wave filter are grown
Degree, and discrete signal in the window of definite wave filter；
Average and variance determination unit, current local radius is determined for discrete signal in the window according to the wave filter
Corresponding average and variance；
Gaussian function determination unit, for determining the height of discrete signal in the window of wave filter using the average and variance
This function；
Sum unit, for determining the corresponding Gauss of each discrete signal in the window of wave filter using the Gaussian function
Functional value, and sum to the Gaussian function numerical value of acquisition；
Weight value cell, for utilizing each corresponding Gaussian function numerical value of discrete signal and Gauss in the window of wave filter
The weighted value corresponding with each discrete signal in the window of definite wave filter of functional value；
Noise suppression unit, noise suppressed is carried out for the wave filter using weighted value to the discrete signal of window center
Processing.
Preferably, the local radius determination unit is specifically used for：
When data index value i is less than optimal local radius, then the local radius of wave filter is equal to data index value i；
When data index value i is more than or equal to optimal local radius, and data index value i is less than the number of discrete signal and subtracts
During optimal local radius, then, the local radius of wave filter is equal to optimal local radius；
Otherwise, the local radius of wave filter is equal to the number of discrete signal and subtracts current data index value and subtract 1 again.
Preferably, the length of window for the wave filter that discrete signal determination unit obtains is described equal to twice in the window
The local radius of wave filter adds 1 again.
Preferably, the noise suppression unit is specifically used for the wave filter by the discrete signal in window and corresponding power
Weight values carry out dot-product operation, which is that wave filter exports the noise suppressed result of currency.
Preferably, the local radius determination unit includes：
Simulator and noise signaling module, for adding noise to emulation signal；
Noise elimination module, for carrying out noise suppressed processing to the emulation signal after addition noise, obtains noise cancellation signal；
Mean square error module, for obtaining the mean square error between emulation signal and the noise cancellation signal corresponding to current radius
Difference；
Optimal local radius determining module, for equal between the emulation signal corresponding to current radius and noise cancellation signal
Square error is with the emulation signal corresponding to upper Radius compared with the mean square error between noise cancellation signal；If current radius
Mean square error between corresponding emulation signal and noise cancellation signal is more than or equal to the emulation signal corresponding to upper Radius with disappearing
Mean square error between noise cancellation signal, then current radius is optimal local radius；Otherwise, lower Radius is as current radius, currently
Radius compares the mean square error between emulation signal and the noise cancellation signal corresponding to current radius and upper half as upper Radius
The mean square error between emulation signal and noise cancellation signal corresponding to footpath, until obtaining optimal local radius.
Above-mentioned technical proposal has the advantages that：
1st, adaptive polo placement weights strategy proposed by the invention, enriches weighting filter design method；
2nd, for containing normal distribution noise and impact noise signal, the more conventional average filter of this filter filtering effect
Ripple device and fixed weighting coefficient filter effect are good；
3rd, compared with complicated approach such as shape filtering, particle filters, inventive algorithm is simple, and it is easy to realize.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is attached drawing needed in technology description to be briefly described, it should be apparent that, drawings in the following description are only this
Some embodiments of invention, for those of ordinary skill in the art, without creative efforts, can be with
Other attached drawings are obtained according to these attached drawings.
Fig. 1 is method flow schematic diagram provided in an embodiment of the present invention；
Fig. 2 is schematic device provided in an embodiment of the present invention；
Fig. 3 is system schematic provided in an embodiment of the present invention；
Fig. 4 is local radius and the graph of relation of mean square error；
Fig. 5 is the muting signal waveforms of the present embodiment；
Fig. 6 is the signal waveforms of the Noise of the present embodiment；
Fig. 7 is the signal waveforms after the de-noising of the present embodiment.
Embodiment
Below in conjunction with the attached drawing in the embodiment of the present invention, the technical solution in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other without making creative work
Embodiment, belongs to the scope of protection of the invention.
One skilled in the art will appreciate that embodiments of the present invention can be implemented as a kind of system, device, equipment, method
Or computer program product.Therefore, the disclosure can be implemented as following form, i.e.,：Complete hardware, complete software
(including firmware, resident software, microcode etc.), or the form that hardware and software combines.
According to the embodiment of the present invention, it is proposed that a kind of noise suppressing method and device based on Gaussian Profile weighting.
Herein, it is to be understood that in involved term：
Gaussian function：Gaussian function is a kind of widely used function in mathematical statistics.It is defined as follows, if stochastic variable
It is μ, standard variance σ that X, which obeys a mathematic expectaion,2Gaussian Profile, be denoted as：X~N (μ, σ2), then its probability density function
For：
In addition, any number of elements in attached drawing is used to example and unrestricted, and any name is only used for distinguishing,
Without any restrictions implication.
Below with reference to some representative embodiments of the present invention, the principle of the present invention and spirit are explained in detail.
Summary of the invention
The technical program is related to a kind of equipment, method and apparatus, the technical program assume in present filter window from
Scattered signal Normal Distribution, the average and variance of the discrete signal in window, determines Gaussian function expression formula；Utilize the Gauss
Function expression determines the corresponding Gaussian function numerical value of discrete signal in window, and Gaussian function numerical value is summed.Will be discrete in window
The sum of the corresponding each Gaussian function numerical value of signal divided by Gaussian function numerical value, so that it is determined that the corresponding weight of discrete signal in window
Value.Discrete signal in window is made into point multiplication operation with corresponding weighted value, its result is the noise cancellation signal of wave filter output.Should
Technical solution has the effect realized Gaussian noise and suppressed while impact noise concurrently.
After the basic principle of the present invention is described, lower mask body introduces the various non-limiting embodiment party of the present invention
Formula.
Application scenarios overview
In hydroelectric power plant when monitoring rotating machinery shaft displacement, due to artificially causing axis surface to contain when manufacturing, install or debugging
There are projection or depression and make the noise for containing similar impact component in signal；Big quantity sensor is arranged in condition monitoring system, its
In it is many be used for monitoring the sensors such as Generator Vibration, throw, partial discharge be installed under strong magnetic environment, being easy to be interfered makes
Obtain signal and produce distortion.In addition, the data packet of sampled signal is caused continuous when passing through transmission of network, due to electromagnetic interference
It is distorted in sampled signal and produces abnormal impact etc. in the signal.With the technical program, sensor signal is enabled to
Effective information is reduced, and suppresses the interference signal under strong magnetic environment, the electromagnetic interference signal in network transmission and class
Like the noise of impact component.The vibration of generator, throw, partial discharge etc. in accurate monitoring system.
Illustrative methods
With reference to application scenarios, the method for exemplary embodiment of the invention is introduced with reference to figure 1.
Understand spirit and principles of the present invention it should be noted that above application scene is for only for ease of and show, this
The embodiment of invention is unrestricted in this regard.On the contrary, embodiments of the present invention can be applied to it is applicable any
Scene.
It is method flow schematic diagram provided in an embodiment of the present invention referring to Fig. 1.As shown in the figure, the step of noise suppressing method
Suddenly include：
Step 101)：Determine the local radius of wave filter；
In a step 101, the local radius of wave filter has three kinds of situations.Discrete signal x (0) using length as N ... x
(N-1) exemplified by.Respectively：When data index value i is less than optimal local radius r0When, then the local radius r of wave filter is equal to number
According to index value i；
When data index value i is more than or equal to optimal local radius r0, and data index value i is less than the number N of discrete signal
Subtract optimal local radius r0When, then the local radius r of wave filter is equal to optimal local radius r0；
Otherwise, the local radius r of wave filter is equal to the number N of discrete signal and subtracts current data index value i and subtract 1 again.
Optimal local radius r0Determine that step includes：
Noise is added to emulation signal；
Noise suppressed processing is carried out to the emulation signal after addition noise, obtains noise cancellation signal；
Obtain the mean square error between emulation signal and the noise cancellation signal corresponding to current radius；
To corresponding to the mean square error between the emulation signal corresponding to current radius and noise cancellation signal and upper Radius
Signal is emulated compared with the mean square error between noise cancellation signal；If the emulation signal corresponding to current radius is believed with de-noising
Mean square error between number is more than or equal to the mean square error between emulation signal and noise cancellation signal corresponding to upper Radius, then when
Preceding radius is optimal local radius；Otherwise, lower Radius is relatively more current as upper Radius as current radius, current radius
The mean square error emulated between signal and noise cancellation signal and the emulation signal corresponding to upper Radius and de-noising corresponding to radius
Mean square error between signal, until obtaining optimal local radius.
Step 102)：The length of window of wave filter is determined according to the local radius of the wave filter, and determines wave filter
Discrete signal in window；
In a step 102, the local radius r of the wave filter of the length of window of the wave filter equal to twice adds 1 again.
Step 103)：The corresponding average of current local radius and side are determined according to random signal in the window of the wave filter
Difference；
Step 104)：The Gaussian function of discrete signal in the window of wave filter is determined using the average and variance；
Step 105)：The corresponding Gaussian function of each discrete signal in the window of wave filter is determined using the Gaussian function
Value, and sum to the Gaussian function numerical value of acquisition；
Step 106)：Utilize the corresponding Gaussian function numerical value of each discrete signal and Gaussian function numerical value in the window of wave filter
Weighted value corresponding with each discrete signal in the window of definite wave filter；
Step 107)：The wave filter carries out noise suppressed processing using weighted value to the discrete signal of window center.
In step 107, the discrete signal in window is carried out dot-product operation by the wave filter with corresponding weighted value, should
Operation result is that wave filter exports the noise suppressed result of currency.
It should be noted that although in the accompanying drawings with the operation of particular order the invention has been described method, still, this is not required that
Or imply and must perform these operations according to the particular order, or the operation having to carry out shown in whole could realize the phase
The result of prestige.Additionally or alternatively, it is convenient to omit multiple steps are merged into a step and performed by some steps, and/or will
One step is decomposed into execution of multiple steps.
Exemplary means
After the method for exemplary embodiment of the invention is described, next, with reference to figure 2 respectively to example of the present invention
The device of property embodiment is introduced.
As shown in Fig. 2, it is device block diagram provided in an embodiment of the present invention.Noise Suppression Device includes：
Local radius determination unit 201, for determining the local radius of wave filter；
The local radius determination unit 201 is specifically used for：
When data index value i is less than optimal local radius, then the local radius of wave filter is equal to data index value i；
When data index value i is more than or equal to optimal local radius, and data index value i is less than the number of discrete signal and subtracts
During optimal local radius, then the local radius of wave filter is equal to optimal local radius；
Otherwise, the local radius of wave filter is equal to the number of discrete signal and subtracts current data index value and subtract 1 again.
Further, include for the optimal local radius that foregoing is directed to, the local radius determination unit：
Simulator and noise signaling module, for adding noise to emulation signal；
Noise elimination module, for carrying out noise suppressed processing to the emulation signal after addition noise, obtains noise cancellation signal；
Mean square error module, for obtaining the mean square error between emulation signal and the noise cancellation signal corresponding to current radius
Difference；
Optimal local radius determining module, for equal between the emulation signal corresponding to current radius and noise cancellation signal
Square error is with the emulation signal corresponding to upper Radius compared with the mean square error between noise cancellation signal；If current radius
Mean square error between corresponding emulation signal and noise cancellation signal is more than or equal to the emulation signal corresponding to upper Radius with disappearing
Mean square error between noise cancellation signal, then current radius is optimal local radius；Otherwise, lower Radius is as current radius, currently
Radius compares the mean square error between emulation signal and the noise cancellation signal corresponding to current radius and upper half as upper Radius
The mean square error between emulation signal and noise cancellation signal corresponding to footpath, until obtaining optimal local radius.
Discrete signal determination unit 202 in window, for determining the window of wave filter according to the local radius of the wave filter
Mouth length, and discrete signal in the window of definite wave filter；
The length of window for the wave filter that discrete signal determination unit 202 obtains is equal to twice of the filtering in the window
The local radius of device adds 1 again.
Average and variance determination unit 203, current local is determined for discrete signal in the window according to the wave filter
The corresponding average of radius and variance；
Gaussian function determination unit 204, for determining discrete signal in the window of wave filter using the average and variance
Gaussian function；
Sum unit 205, for determining that each discrete signal is corresponding in the window of wave filter using the Gaussian function
Gaussian function numerical value, and sum to the Gaussian function numerical value of acquisition；
Weight value cell 206, for using each corresponding Gaussian function numerical value of discrete signal in the window of wave filter and
The weighted value corresponding with each discrete signal in the window of definite wave filter of Gaussian function numerical value；
Noise suppression unit 207, noise is carried out for the wave filter using weighted value to the discrete signal of window center
Suppression is handled.
Further, the noise suppression unit 207 be specifically used for the wave filter by the discrete signal in window with it is right
The weighted value answered carries out dot-product operation, which is the knot that wave filter handles the discrete signal noise suppressed in window
Fruit.
In addition, although being referred to some units of device in above-detailed, but this division is only not strong
Property processed.In fact, according to the embodiment of the present invention, the feature and function of two or more above-described units can be
Embodied in one unit.Equally, the feature of an above-described unit and function can also be further divided into by multiple
Unit embodies.
Example devices
Based on above-mentioned example apparatus and method, the present embodiment also proposes a kind of equipment, as shown in Figure 3.The system is used for
Noise suppressed；Including：
Memory a, for storing request instruction；
Processor b, it is coupled with the memory, which, which is configured as performing, is stored in asking in the memory
Instruction is asked, wherein, the application program that the processor is configured is used for：
Determine the local radius of wave filter；
The length of window of wave filter is determined according to the local radius of the wave filter, and it is discrete in the window of definite wave filter
Signal；
The corresponding average of current local radius and variance are determined according to discrete signal in the window of the wave filter；
The Gaussian function of discrete signal in the window of wave filter is determined using the average and variance；
The corresponding Gaussian function numerical value of each discrete signal in the window of wave filter is determined using the Gaussian function, and to obtaining
The Gaussian function numerical value summation obtained；
Using the corresponding Gaussian function numerical value of each discrete signal in the window of wave filter and Gaussian function numerical value and determine
The corresponding weighted value of each discrete signal in the window of wave filter；
The wave filter carries out noise suppressed processing using weighted value to the discrete signal of window center.
The embodiment of the present invention also provides a kind of computer-readable program, wherein when performing described program in the electronic device
When, described program causes computer to perform the noise suppression based on Gaussian Profile weighting as described in Figure 1 in the electronic equipment
The method of system.
The embodiment of the present invention also provides a kind of storage medium for being stored with computer-readable program, wherein the computer can
Reader causes the method that computer performs the noise suppressed based on Gaussian Profile weighting as described in Figure 1 in the electronic device.
Embodiment
In order to more intuitively describe the features of the present invention and operation principle, below in conjunction with a practice field
Scape describes.
As shown in figure 4, it is local radius and the graph of relation of mean square error.In the present embodiment, selection emulation signal
Local radius with least mean-square error is as optimal local radius.In the present embodiment, optimal local radius is 5.
As shown in figure 5, the muting signal waveforms for the present embodiment.As shown in fig. 6, the Noise for the present embodiment
Signal waveforms.As shown in fig. 7, it is the signal waveforms after the de-noising of the present embodiment.In Figure 5, the not name of Noise
For the signal of " bumps ".By operation, the noise and impact noise of normal distribution, 5 He of comparison diagram are added in bumps signals
Fig. 6 has found that the waveform of signal changes.It is prerequisite using optimal local radius as 5, using the technical program to Fig. 6's
Signal carries out noise suppressed processing, obtains the signal waveform shown in Fig. 7.Comparison diagram 5 and Fig. 7 are it can be found that signal in two figures
Waveform is basically identical, it can be seen that it is effectively suppressed to the noise of the normal distribution of addition with impact noise, noise suppressed
For oscillogram afterwards with actual very close, noise suppression effect is fine.
Above embodiment, has carried out further specifically the purpose of the present invention, technical solution and beneficial effect
It is bright, it should be understood that these are only the embodiment of the present invention, the protection model being not intended to limit the present invention
Enclose, within the spirit and principles of the invention, any modification, equivalent substitution, improvement and etc. done, should be included in the present invention
Protection domain within.
Claims (8)
- A kind of 1. noise suppressing method based on Gaussian Profile weighting, it is characterised in that including：Determine the local radius of wave filter；The length of window of wave filter, and discrete letter in the window of definite wave filter are determined according to the local radius of the wave filter Number；The corresponding average of current local radius and variance are determined according to discrete signal in the window of the wave filter；The Gaussian function of discrete signal in the window of wave filter is determined using the average and variance；The corresponding Gaussian function numerical value of each discrete signal in the window of wave filter is determined using the Gaussian function, and to acquisition Gaussian function numerical value is summed；Using the corresponding Gaussian function numerical value of each discrete signal in the window of wave filter and Gaussian function numerical value and determine filtering The corresponding weighted value of each discrete signal in the window of device；Discrete signal in window is carried out dot-product operation by the wave filter with corresponding weighted value, which is wave filter The noise suppressed result of currency is exported.
- 2. the method as described in claim 1, it is characterised in that the step of local radius of the definite wave filter includes：When data index value i is less than optimal local radius, then the local radius of wave filter is equal to data index value i；When data index value i is more than or equal to optimal local radius, and data index value i be less than discrete signal number subtract it is optimal During local radius, then, the local radius of wave filter is equal to optimal local radius；Otherwise, the local radius of wave filter is equal to the number of discrete signal and subtracts current data index value and subtract 1 again.
- 3. method as claimed in claim 1 or 2, it is characterised in that the length of window of the wave filter is described equal to twice The local radius of wave filter adds 1 again.
- 4. method as claimed in claim 2, it is characterised in that the optimal local radius obtaining step includes：Noise is added to emulation signal；Noise suppressed processing is carried out to the emulation signal after addition noise, obtains noise cancellation signal；Obtain the mean square error between emulation signal and the noise cancellation signal corresponding to current radius；To the mean square error between the emulation signal corresponding to current radius and noise cancellation signal and the emulation corresponding to upper Radius Compared with mean square error between signal and noise cancellation signal；If emulation signal corresponding to current radius and noise cancellation signal it Between mean square error be more than or equal to upper Radius corresponding to emulation signal and noise cancellation signal between mean square error, then work as first half Footpath is optimal local radius；Otherwise, lower Radius compares current radius as current radius, current radius as upper Radius Mean square error between corresponding emulation signal and noise cancellation signal and the emulation signal and noise cancellation signal corresponding to upper Radius Between mean square error, until obtaining optimal local radius.
- A kind of 5. Noise Suppression Device based on Gaussian Profile weighting, it is characterised in that including：Local radius determination unit, for determining the local radius of wave filter；Discrete signal determination unit in window, for determining the length of window of wave filter according to the local radius of the wave filter, And discrete signal in the window of definite wave filter；Average and variance determination unit, determine that current local radius corresponds to for discrete signal in the window according to the wave filter Average and variance；Gaussian function determination unit, for determining the Gaussian function of discrete signal in the window of wave filter using the average and variance Number；Sum unit, for determining the corresponding Gaussian function of each discrete signal in the window of wave filter using the Gaussian function Value, and sum to the Gaussian function numerical value of acquisition；Weight value cell, for utilizing each corresponding Gaussian function numerical value of discrete signal and Gaussian function in the window of wave filter The weighted value corresponding with each discrete signal in the window of definite wave filter of value；Noise suppression unit, dot-product operation is carried out for the wave filter by the discrete signal in window with corresponding weighted value, The operation result is that wave filter exports the noise suppressed result of currency.
- 6. device as claimed in claim 5, it is characterised in that the local radius determination unit is specifically used for：When data index value i is less than optimal local radius, then the local radius of wave filter is equal to data index value i；When data index value i is more than or equal to optimal local radius, and data index value i be less than discrete signal number subtract it is optimal During local radius, then, the local radius of wave filter is equal to optimal local radius；Otherwise, the local radius of wave filter is equal to the number of discrete signal and subtracts current data index value and subtract 1 again.
- 7. the device as described in claim 5 or 6, it is characterised in that the filter that discrete signal determination unit obtains in the window The local radius of the wave filter of the length of window of ripple device equal to twice adds 1 again.
- 8. device as claimed in claim 6, it is characterised in that the local radius determination unit includes：Simulator and noise signaling module, for adding noise to emulation signal；Noise elimination module, for carrying out noise suppressed processing to the emulation signal after addition noise, obtains noise cancellation signal；Mean square error module, for obtaining the mean square error between emulation signal and the noise cancellation signal corresponding to current radius；Optimal local radius determining module, for the mean square error between the emulation signal corresponding to current radius and noise cancellation signal The poor emulation signal with corresponding to upper Radius is compared with the mean square error between noise cancellation signal；If current radius institute is right The emulation signal that mean square error between the emulation signal and noise cancellation signal answered is more than or equal to corresponding to upper Radius is believed with de-noising Mean square error between number, then current radius is optimal local radius；Otherwise, lower Radius is as current radius, current radius As upper Radius, compare the mean square error between emulation signal and the noise cancellation signal corresponding to current radius and upper Radius institute Mean square error between corresponding emulation signal and noise cancellation signal, until obtaining optimal local radius.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610169563.2A CN105812068B (en) | 2016-03-23 | 2016-03-23 | A kind of noise suppressing method and device based on Gaussian Profile weighting |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610169563.2A CN105812068B (en) | 2016-03-23 | 2016-03-23 | A kind of noise suppressing method and device based on Gaussian Profile weighting |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105812068A CN105812068A (en) | 2016-07-27 |
| CN105812068B true CN105812068B (en) | 2018-05-04 |
Family
ID=56454735
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610169563.2A Active CN105812068B (en) | 2016-03-23 | 2016-03-23 | A kind of noise suppressing method and device based on Gaussian Profile weighting |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105812068B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1459761A (en) * | 2002-05-24 | 2003-12-03 | 清华大学 | Character identification technique based on Gabor filter set |
| CN1905006A (en) * | 2005-07-27 | 2007-01-31 | 日本电气株式会社 | Noise suppression system, method and program |
| CN101149787A (en) * | 2006-09-20 | 2008-03-26 | 中国科学院自动化研究所 | Fingerprint synthesis method based on orientation field model and Gabor filter |
| CN101464998A (en) * | 2009-01-15 | 2009-06-24 | 浙江大学 | Non-gauss veins noise smooth filtering method for textile industry |
| CN102567973A (en) * | 2012-01-06 | 2012-07-11 | 西安电子科技大学 | Image denoising method based on improved shape self-adaptive window |
| CN102855612A (en) * | 2012-09-05 | 2013-01-02 | 天津光电高斯通信工程技术有限公司 | Adaptive enhancement algorithm of CCD (charged coupled device) image based on linear gray array |
| CN103942759A (en) * | 2014-04-08 | 2014-07-23 | 武汉烽火众智数字技术有限责任公司 | Three-dimensional noise reduction method and device based on Gaussian background model in fixed scene |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| ATE395669T1 (en) * | 2004-11-24 | 2008-05-15 | Koninkl Philips Electronics Nv | TIME FILTERING WITH MULTIPLE FEATURES TO IMPROVE THE STRUCTURE OF NOISED IMAGES |
-
2016
- 2016-03-23 CN CN201610169563.2A patent/CN105812068B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1459761A (en) * | 2002-05-24 | 2003-12-03 | 清华大学 | Character identification technique based on Gabor filter set |
| CN1905006A (en) * | 2005-07-27 | 2007-01-31 | 日本电气株式会社 | Noise suppression system, method and program |
| CN101149787A (en) * | 2006-09-20 | 2008-03-26 | 中国科学院自动化研究所 | Fingerprint synthesis method based on orientation field model and Gabor filter |
| CN101464998A (en) * | 2009-01-15 | 2009-06-24 | 浙江大学 | Non-gauss veins noise smooth filtering method for textile industry |
| CN102567973A (en) * | 2012-01-06 | 2012-07-11 | 西安电子科技大学 | Image denoising method based on improved shape self-adaptive window |
| CN102855612A (en) * | 2012-09-05 | 2013-01-02 | 天津光电高斯通信工程技术有限公司 | Adaptive enhancement algorithm of CCD (charged coupled device) image based on linear gray array |
| CN103942759A (en) * | 2014-04-08 | 2014-07-23 | 武汉烽火众智数字技术有限责任公司 | Three-dimensional noise reduction method and device based on Gaussian background model in fixed scene |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105812068A (en) | 2016-07-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100550666C (en) | Narrow band interference is eliminated in the spread spectrum system method, device | |
| CN104459398B (en) | A kind of quality of power supply of use Two-dimensional morphology noise reduction is combined disturbance identification method | |
| CN108764073B (en) | Acceleration noise filtering and integrating method combined with spectrum energy form fitting | |
| Pavlenko et al. | Identification of systems using Volterra model in time and frequency domain | |
| CN106157967A (en) | Impulse noise mitigation | |
| CN108445313B (en) | Comprehensive prediction method for electromagnetic radiation effect of frequency equipment and terminal equipment | |
| CN107870316A (en) | A kind of TDOA localization methods calculated based on the time difference, apparatus and system | |
| JPWO2006019196A1 (en) | Electromagnetic interference reduction amount calculation method, electromagnetic interference reduction amount calculation device, calculation program, and electronic circuit | |
| CN105785324A (en) | MGCSTFT-based chirp signal parameter estimation method | |
| CN113283297A (en) | Method for estimating partial discharge signal time based on singular value decomposition and improved wavelet packet | |
| CN110081967A (en) | Mechanical oscillation signal Threshold Noise Reduction Methods based on spectrogram wavelet transformation | |
| Abdullah et al. | Power quality analysis using bilinear time-frequency distributions | |
| CN113608023A (en) | Sampling frequency self-adaptive harmonic detection method and system | |
| CN114325598A (en) | Signal denoising method and device | |
| CN114358093A (en) | Method and equipment for detecting partial discharge in power equipment | |
| Tian et al. | Early fault feature extraction for rolling bearings using adaptive variational mode decomposition with noise suppression and fast spectral correlation | |
| CN107991531A (en) | Power grid harmonic wave detection method based on variation mode decomposition and removal noise | |
| JP2011527160A (en) | Dynamic filtering for adjacent channel interference suppression | |
| Wunderlich et al. | Data-driven modeling of noise time series with convolutional generative adversarial networks∗ | |
| CN113054949B (en) | Filtering method, device and equipment for water hammer pressure wave signal | |
| CN105812068B (en) | A kind of noise suppressing method and device based on Gaussian Profile weighting | |
| CN112800863A (en) | Time-frequency feature extraction method and system for communication signal modulation pattern recognition | |
| CN107481732A (en) | Noise reduction method and device in spoken language evaluation and terminal equipment | |
| CN110287853B (en) | Transient signal denoising method based on wavelet decomposition | |
| CN105652166B (en) | A kind of Weighted Threshold wavelet de-noising method for partial discharge on-line monitoring |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |