CN101662679A - Method and device for filtering waves - Google Patents
Method and device for filtering waves Download PDFInfo
- Publication number
- CN101662679A CN101662679A CN 200810146860 CN200810146860A CN101662679A CN 101662679 A CN101662679 A CN 101662679A CN 200810146860 CN200810146860 CN 200810146860 CN 200810146860 A CN200810146860 A CN 200810146860A CN 101662679 A CN101662679 A CN 101662679A
- Authority
- CN
- China
- Prior art keywords
- component
- matching degree
- luminance component
- chromatic
- filter factor
- 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.)
- Granted
Links
Images
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention embodiment discloses a method and a device for filtering waves. The method comprises the following steps: adjusting the initial filter coefficients of a luminance component and/or a chrominance component according to a combination of matching scores of the luminance component and the chrominance component; and acquiring a total filter coefficient by using the adjusted filter coefficient of the luminance component and the adjusted filter coefficient of the chrominance component. With method and the device, the realization complexity is reduced on the premise of guaranteeing the filtering effect and the filtering distortion caused by filtering under the condition of a large difference between the matching scores of the luminance component and the chrominance component is removed.
Description
Technical field
The present invention relates to field of network communication, relate in particular to a kind of method and apparatus of filtering.
Background technology
MCTF (Motion Compensated Temporal Filter, motion compensated temporal filter) framework schematic diagram, as shown in Figure 1, the realization principle of this MCTF is as follows: the piece that present frame is divided into a plurality of M*N, carry out estimation before present frame or in the reference frame afterwards, seek match block for the piece that will carry out filtering in the present frame, in current block and match block, carry out time-domain filtering then.Wherein each piece only uses a motion vector, and just all pixels in piece are used same motion vector, are that each piece in the frame finds the process of optimum movement vector to be estimation.
At present, the implementation of a kind of known MCTF is: in whole filtering motion compensated temporal filter process, the handling process of luminance component and chromatic component is separate, each has independently motion search module, filter factor computing module and filtration module.
The inventor finds that there is following problem at least in prior art in realizing process of the present invention:
The handling process of luminance component of the prior art and chromatic component is separate, each have independently motion search module, because motion search can expend a large amount of resources, so simultaneously luminance component and chromatic component are carried out the complexity that motion search can increase scheme, improve and realize the cost height, and because the handling process of luminance component and chromatic component is fully independently, when luminance component filter factor and chromatic component filter factor differed big, the distortion of tone reversal may appear.
Summary of the invention
The embodiment of the invention provides a kind of method and apparatus of filtering, is guaranteeing to reduce the complexity that realizes under the prerequisite of filter effect, and eliminates because there is the distortion that causes during filtering than under the big-difference situation in the matching degree of the matching degree of luminance component and chromatic component.
The embodiment of the invention provides a kind of method of filtering, comprising:
According to the combination of the matching degree of the matching degree of luminance component and chromatic component, adjust the initial filter coefficients of described luminance component and/or chromatic component;
Obtain total filter factor by the filter factor that passes through the described luminance component of adjusting and the filter factor of chromatic component.
The embodiment of the invention also provides a kind of device of filtering, comprising:
The filter factor adjusting module is used for the combination according to the matching degree of the matching degree of luminance component and chromatic component, adjusts the initial filter coefficients of described luminance component and/or chromatic component;
Total filter factor acquisition module is used for obtaining total filter factor by the filter factor that passes through the described luminance component of adjusting and the filter factor of chromatic component.
Compared with prior art, the embodiment of the invention has the following advantages:
The method and apparatus of the filtering that provides by the embodiment of the invention, can obtain the motion vector of chromatic component according to the motion vector of luminance component, reduce the complexity that realizes, simultaneously according to the combination of the matching degree of the matching degree of luminance component and chromatic component, adjust the initial filter coefficients of described luminance component and/or chromatic component, eliminated because there is the image fault that causes than big-difference in the matching degree of the matching degree of luminance component and chromatic component.
Description of drawings
In order to be illustrated more clearly in the embodiment of the invention or technical scheme of the prior art, to do to introduce simply to the accompanying drawing of required use in embodiment or the description of the Prior Art below, apparently, accompanying drawing in describing below only is some embodiments of the present invention, for those of ordinary skills, under the prerequisite of not paying creative work, can also obtain other accompanying drawing according to these accompanying drawings.
Fig. 1 is the realization principle schematic of MCTF in the prior art;
Fig. 2 is the method flow diagram of filtering in one embodiment of the invention;
Fig. 3 is the method flow diagram of filtering in another embodiment of the present invention;
Fig. 4 is the computational process schematic diagram of matching degree in another embodiment of the present invention;
Fig. 5 is the apparatus structure schematic diagram of filtering in the embodiment of the invention.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the invention, the technical scheme in the embodiment of the invention is clearly and completely described, obviously, described embodiment only is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills belong to the scope of protection of the invention not making the every other embodiment that is obtained under the creative work prerequisite.
The embodiment of the invention provides a kind of method and apparatus of filtering, is guaranteeing to reduce the complexity that realizes under the prerequisite of filter effect, and eliminates because there is the distortion that causes during filtering than under the big-difference situation in the matching degree of the matching degree of luminance component and chromatic component.
The method and apparatus of the filtering that the embodiment of the invention is provided below in conjunction with the drawings and specific embodiments is elaborated.
The embodiment of the invention provides a kind of method of filtering, as shown in Figure 2, comprising:
Step S201, according to the combination of the matching degree of the matching degree of luminance component and chromatic component, adjust the initial filter coefficients of luminance component and/or chromatic component.Concrete, obtain the match block of luminance component by motion search, obtain the matching degree and the motion vector of luminance component according to this match block.By the form of the digital video signal relation of luminance component and chromatic component as can be known, the motion vector of adjusting luminance component according to this relation obtains the motion vector of chromatic component.Owing to do not need chromatic component is carried out motion search, reduced the complexity when this method realizes.
When the matching degree of three components all is when not matching fully, the filter coefficient alpha l when just three components have all been selected weak filtering simultaneously, the initial filter coefficients of three components does not need to adjust in this case; When the matching degree of three components all is when mating fully, the filter coefficient alpha h when just three components have all been selected strong filtering simultaneously, the initial filter coefficients of three components does not need to adjust in this case; In three components, occur not matching fully without any one-component, when three components also not occurring and all being the situation of mating fully, the initial filter coefficients of the component that matching degree is bigger need be adjusted to the initial filter coefficients direction of the component of matching degree minimum.
Step S202, by obtaining total filter factor through the filter factor of the luminance component adjusted and the filter factor of chromatic component.After according to total filter factor luminance component and chromatic component being carried out filtering, according to synthetic this filtered luminance component of form and the chromatic component of the digital video signal of input and export.
By above-mentioned steps, only need carry out motion search to luminance component, reduced the complexity that realizes, reduced the realization cost; Because the initial filter coefficients of luminance component and chromatic component is adjusted, preferably resolve the distorted signals problem simultaneously.
Another embodiment of the present invention provides a kind of method of filtering, as shown in Figure 3, comprising:
Step S301, digital video signal is separated into luminance component and chromatic component.
Step S302, luminance component is carried out motion search, obtain the match block of filtering image piece in reference frame in the present frame.Concrete, can adopt fast search algorithm when utilizing the motion search best matching blocks, also can adopt full-search algorithm, generally speaking, motion estimation process adopts " least energy method " to search the match block of filtering image piece in reference frame in the present frame, and obtains best motion vector.Should " least energy method " refer to search for the image block of energy minimum in the residual block, the energy of this residual block can use MSE (Mean Square Error, least mean-square error) or SSD (sum of square error, residual sum of squares (RSS)) equivalence to characterize.
Step S303, according to the matching degree of filtering image piece in the match block of luminance component and the present frame, determine initial filter coefficients shared weight in total filter factor of the motion vector and the luminance component of luminance component.The MCTF filtering is the process of respectively luminance component and chromatic component being carried out filtering, so the matching degree of luminance component can only partly determine final total filter factor, this total filter factor also matching degree with chromatic component is relevant.
Step S304, the motion vector of luminance component is adjusted, obtained being fit to the motion vector of chromatic component according to the form of digital video signal.Digital video signal has multiple form, (as the 4:4:4 form in ITURecommendation 601/656 standard, 4:2:2 form, 4:2:0 form etc.), according to the relation of chromatic component and luminance component in the form of data video signal and the motion vector of luminance component, adjust the motion vector of the chromatic component that obtains being fit to.
Step S305, according to the matching degree of the motion vector computation chromatic component of chromatic component.
Step S306, calculate initial filter coefficients shared weight in total filter factor of chromatic component according to the matching degree of chromatic component.
Step S307, according to the initial filter coefficients of matching degree combination calculation/each component of adjustment of luminance component and chromatic component.
Step S308, according to initial filter coefficients shared weight in total filter factor of the initial filter coefficients of luminance component shared weight and chromatic component in total filter factor, calculate total filter factor.
Step S309, luminance component and chromatic component are carried out time-domain filtering respectively.This luminance component comprises the luminance component of present frame and the luminance component of reference frame, and when carrying out filtering, the parameter of input also comprises the motion vector and the filter factor of the luminance component that obtains by motion search.This chromatic component filtering is identical with luminance component.
Output and digital video signal behind step S310, synthetic luminance component and the chromatic component, the form of this digital video signal is identical with the digital video signal of input.Luminance component and chromatic component are through behind the motion compensated temporal filter, with the identical vision signal of format digital video signal that Y, U, V component synthesize and import, output then.
In the foregoing description, the matching degree of luminance component Y is identical with the computational methods of the matching degree of chromatic component U, V among step S302 and the step S305.As shown in Figure 4, computational process for matching degree, the residual energy of residual block is mainly used in the calculating of this matching degree, this residual energy can be used a kind of expression the in the following parameter: MSE, SSD, SAD (Sum Of Absolute Difference, residual absolute value and) or MAD (Mean Absolute Deviation, mean absolute deviation) etc., describe with the SSD value here.The trunnion axis of this figure is represented the SSD value, as seen from the figure, the preset value of two SSD: SSD_th1 and SSD_th2 is arranged on the trunnion axis.If the SSD value of calculative component is less than or equal to SSD_th1, then the filtering image piece of present frame and the reference block in the reference frame mate fully, and matching degree is represented with Upper_th; If the SSD value of calculative component is during greater than SSD_th2, then the filtering image piece of present frame and the reference block in the reference frame do not match fully, and all Lower_th of matching degree represent; When if the SSD value of the component that calculates is less than or equal to SSD_th2 greater than SSD_th1, then matching degree between fully the coupling and do not match fully between, matching degree and the SSD value relation of being inversely proportional in this interval, promptly be that the more little matching degree of SSD value is high more, the big more matching degree of SSD value is low more, can use following formula when calculating matching degree:
Matching degree σ=[(Upper_th-Lower_th)/(SSD_th1-SSD_th2)] * (SSD-SSD_th1)+Upper_th
In the foregoing description, total filter factor determines jointly that by the filter factor of luminance component and chromatic component computational process need be used the matching degree of luminance component and chromatic component, and the filter factor of luminance component and chromatic component is relatively independent.
The initial filter coefficients of luminance component and chromatic component can be divided into three class:
Strong filtering: when this filter factor is used in expression, be the strongest to the filtering strength of this component;
Weak filtering: when this filter factor is used in expression, be the most weak to the filtering strength of this component;
According to the matching degree calculating filter coefficient: adopt matching degree formula calculating filter coefficient, the filtering strength of this component is between strong filtering and weak filtering.
For convenience of explanation, the filter factor the when filter factor in the time of can using α h to represent strong filtering, α l are represented weak filtering, wherein α h, α l are predetermined constants; α represents that according to matching degree formula calculating filter coefficient this filter factor changes with matching degree is different, and the span of α is α l＜=α＜=α h.It is more little to define filter factor herein, and filtering strength is big more, after filter factor normalization, and 0＜=α＜=1, then the computing formula of α is as follows:
α＝(αl-αh)/(Lower_th-Upper_th)*(σ-Lower_th)+αl
In the above-mentioned formula, the matching degree when Lower_th represents that the filtering image piece of present frame and the reference block in the reference frame do not match fully, the matching degree when Upper_th represents that the filtering image piece of present frame and the reference block in the reference frame mate fully.
Below for the aspect explanation, the digital video signal that uses ITU Recommendation 601/656 form is the adjustment process of the filter factor of example explanation brightness and chromatic component and brightness, chromatic component filtering, wherein Y represents luminance component, and U represents the C in the chromatic component
BComponent, V are represented the C in the chromatic component
RComponent.But the present invention is not limited only to handle this format digital video signal.
Because filter factor and matching degrees luminance Y component, colourity U component, colourity V component three components relevant with matching degree respectively have fully coupling, do not match fully and three kinds of situations are mated in the centre, can make up the initial filter coefficients of calculating/adjust each component according to the matching degree of luminance Y component, colourity U component and V component, the calculating/adjustment process of each filter factor is as follows:
(1) calculate the initial filter factor of each component according to the matching degree of luminance Y component, colourity U component and colourity V component, the initial filter coefficients of each component shared weight in its final filter factor also depends on the matching degree of other components.
Initial filter coefficients is calculated principle:
When the matching degree of certain component is when mating fully, its initial filter coefficients is defined as α h; When the matching degree of certain component is when not matching fully, its initial filter coefficients is defined as α l; When the matching degree of certain component was not in coupling fully and between matching fully, its initial filter coefficients need be calculated according to matching degree.
(2) adjust the initial filter coefficients of each component according to the matching degree of luminance Y component, colourity U component and colourity V component, just the final filter factor of each component is relevant with the matching degree factor of other two components.A kind of adjustment initial filter coefficients method is provided below: the initial filter coefficients (being assumed to be α i) with the component of matching degree minimum in three components is a benchmark, the initial filter coefficients of other two components that matching degree is bigger (supposing to be respectively α j and α k) is adjusted to the initial filter coefficients direction of the component of matching degree minimum according to certain rule, and the combination of the matching degree of each component is relevant in this rule and the actual conditions.
According to definition above: filter factor is more little, and filtering strength is big more, should be that the initial filter coefficients (α j and α k) of two components that matching degree is bigger increases according to certain rule so.The final filter factor of supposing above-mentioned three components is respectively α ' i, α ' j and α ' k, so according to following equation is arranged after the mentioned above principle adjustment:
α′i＝αi
α ' j=(1-k1) * α j+k1 * α i k1 is a constant, and general span is 0～1
α ' k=(1-k2) * α k+k2 * α i k2 is a constant, and general span is 0～1
Illustrate the adjustment process of the filter factor of luminance Y component, colourity U component and three components of V component below with form, but in the present embodiment for example be mainly used in and explain the situation, the method that the embodiment of the invention provides not only is confined to cited situation in the practice process.
Table 1
| ??αy | ??αu | ??αv | |
| ??match | ??0 | ??0 | ??0 |
| ??Mid-match | ??0 | ??0 | ??0 |
| ??No-match | ??1 | ??1 | ??1 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
As shown in table 1, when the matching degree of three components all is when not matching fully, the filter coefficient alpha l when just three components have all been selected weak filtering simultaneously, the initial filter coefficients of three components does not need to adjust in this case.
Table 2
| ??αy | ??αu | ??αv | |
| ??match | ??1 | ??1 | ??1 |
| ??Mid-match | ??0 | ??0 | ??0 |
| ??No-match | ??0 | ??0 | ??0 |
| ??Filter-mode | Strong filtering | Strong filtering | Strong filtering |
As shown in table 2, when the matching degree of three components all is when mating fully, the filter coefficient alpha h when just three components have all been selected strong filtering simultaneously, the initial filter coefficients of three components does not need to adjust in this case.
Table 3
| ??αy | ??αu | ??αv | |
| ??Match | ??0 | ??0 | ??0 |
| ??Mid-match | ??0 | ??0 | ??0 |
| ??No-match | ??1 | ??1 | ??1 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??αv | |
| ??Match | ??0 | ??0 | ??0 |
| ??Mid-match | ??0 | ??0 | ??1 |
| ??No-match | ??1 | ??1 | ??0 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??αv | |
| ??Match | ??0 | ??0 | ??1 |
| ??Mid-match | ??0 | ??0 | ??0 |
| ??No-match | ??1 | ??1 | ??0 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??αv | |
| ??Match | ??0 | ??0 | ??0 |
| ??Mid-match | ??0 | ??1 | ??0 |
| ??No-match | ??1 | ??0 | ??1 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??αv | |
| ??Match | ??0 | ??1 | ??0 |
| ??Mid-match | ??0 | ??0 | ??0 |
| ??No-match | ??1 | ??0 | ??1 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??αv | |
| ??Match | ??1 | ??0 | ??0 |
| ??Mid-match | ??0 | ??1 | ??0 |
| ??No-match | ??0 | ??0 | ??1 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??αv | |
| ??Match | ??0 | ??0 | ??0 |
| ??Mid-match | ??1 | ??0 | ??0 |
| ??No-match | ??0 | ??1 | ??1 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??αv | |
| ??Match | ??1 | ??0 | ??0 |
| ??Mid-match | ??0 | ??0 | ??0 |
| ??No-match | ??0 | ??1 | ??1 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??αv | |
| ??Match | ??0 | ??0 | ??0 |
| ??Mid-match | ??0 | ??1 | ??1 |
| ??No-match | ??1 | ??0 | ??0 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??Av | |
| ??Match | ??0 | ??0 | ??1 |
| ??Mid-match | ??0 | ??1 | ??0 |
| ??No-match | ??1 | ??0 | ??0 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??Av | |
| ??Match | ??0 | ??0 | ??0 |
| ??Mid-match | ??1 | ??1 | ??0 |
| ??No-match | ??0 | ??0 | ??1 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??Av | |
| ??Match | ??0 | ??1 | ??0 |
| ??Mid-match | ??1 | ??0 | ??0 |
| ??No-match | ??0 | ??0 | ??1 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??Av | |
| ??Match | ??0 | ??1 | ??1 |
| ??Mid-match | ??0 | ??0 | ??0 |
| ??No-match | ??1 | ??0 | ??0 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??Av | |
| ??Match | ??0 | ??1 | ??0 |
| ??Mid-match | ??0 | ??0 | ??1 |
| ??No-match | ??1 | ??0 | ??0 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??Av | |
| ??Match | ??0 | ??0 | ??0 |
| ??Mid-match | ??1 | ??0 | ??1 |
| ??No-match | ??0 | ??1 | ??0 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??αv | |
| ??Match | ??0 | ??0 | ??1 |
| ??Mid-match | ??1 | ??0 | ??0 |
| ??No-match | ??0 | ??1 | ??0 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??Av | |
| ??Match | ??1 | ??0 | ??1 |
| ??Mid-match | ??0 | ??0 | ??0 |
| ??No-match | ??0 | ??1 | ??0 |
| ??Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ??αy | ??αu | ??Av | |
| ??Match | ??1 | ??0 | ??0 |
| ??Mid-match | ??0 | ??0 | ??1 |
| ??No-match | ??0 | ??1 | ??0 |
| Filter-mode | Weak filtering | Weak filtering | Weak filtering |
| ????αy | ????αu | ??Av | |
| ?Match | ????1 | ????1 | ??0 |
| ?Mid-match | ????0 | ????0 | ??0 |
| ?No-match | ????0 | ????0 | ??1 |
| ?Filter-mode | Weak filtering | Weak filtering | Weak filtering |
As shown in table 3, as long as in the matching degree of three components, there is the matching degree of one-component to be in complete unmatched state, so no matter the matching degree of other two components how, three components are all selected weak filtering, the filter factor of three components all is adjusted into α l.
Table 4
As shown in table 4, in three components, occur not matching fully without any one-component, when three components also not occurring and all being the situation of mating fully, the initial filter coefficients of the component that matching degree is bigger need be adjusted to the initial filter coefficients direction of the component of matching degree minimum.
When among the foregoing description step S309 luminance component being carried out filtering, luminance component comprises the luminance component of present frame and the luminance component of reference frame, this filtering and typical motion compensated temporal filter are similar, can suppose that Yfilter represents filtered luminance component, Yn represents the brightness of present frame, Yn-1 represents the brightness of reference frame, and α y represents the filter factor of luminance component, and filtering is expressed as follows so:
Yfilter＝(1-αy)*Yn-1+αy*Yn
For chromatic component, the parameter that need import when filtering has: chromatic component comprises the chromatic component of present frame and the chromatic component of reference frame; The motion vector of colourity UV component; The filter factor of colourity UV component.The filtering of colourity UV component is similar in luminance component, and its filtering can be used following formulate:
Ufilter＝(1-αu)*Un-1+αu*Un
Vfilter＝(1-αv)*Vn-1+αv*Vn
The embodiment of the invention also provides a kind of device of filtering, as shown in Figure 5, comprising:
Filter factor adjusting module 503 is used for the combination according to the matching degree of the matching degree of luminance component and chromatic component, adjusts the initial filter coefficients of described luminance component and/or chromatic component.This filter factor adjusting module 503 is when adjusting initial filter coefficients, when the matching degree of a luminance component and two chromatic components all is to mate fully or when not matching fully, the initial filter coefficients of three components does not need to adjust; Have the matching degree of one-component to be in complete unmatched state in the matching degree of a luminance component and two chromatic components, the filter factor of three components all is adjusted into weak filter factor; Be to mate fully or do not match fully without any the matching degree of one-component in the matching degree of a luminance component and two chromatic components, the initial filter coefficients of the component that matching degree is bigger need be adjusted to the initial filter coefficients direction of the component of matching degree minimum.
Total filter factor acquisition module 504 is used for obtaining total filter factor by the filter factor that passes through the described luminance component of adjusting and the filter factor of chromatic component.
This device also comprises:
Digital video signal separation module 505 is used for the luminance component of digital video signal is separated with chromatic component;
Brightness match block search module 506 is used at movement compensation process luminance component being carried out motion search, obtains the match block that is complementary with described luminance component;
Brightness movement vector acquisition module 507 is used for the matching degree according to described match block and present frame, obtains the motion vector of described luminance component;
Brightness initial filter coefficients weight acquisition module 508 is used for the matching degree according to described match block and present frame, obtains initial filter coefficients shared weight in total filter factor of described luminance component.
This device also comprises: chroma motion vector acquisition module 501 is used for obtaining according to the motion vector of luminance component the motion vector of chromatic component; Chromaticity match degree acquisition module 502 is used for being obtained by described chroma motion vector the matching degree of described chromatic component.
Colourity initial filter coefficients weight acquisition module 509 is used for obtaining described chromatic component initial filter coefficients in total shared weight of filter factor according to the matching degree of described chromatic component.
This device also comprises:
Brightness time-domain filtering module 510 is used for described luminance component is carried out time-domain filtering according to total filter factor;
Colourity time-domain filtering module 511 is used for described chromatic component is carried out time-domain filtering according to total filter factor;
The method and apparatus of the filtering that provides by the embodiment of the invention, only need carry out motion search to luminance component, with respect to the scheme of simultaneously luminance component Y and chromatic component U, V being carried out motion search, reduced the complexity that realizes, reduced the realization cost of scheme.Because the method that the embodiment of the invention provides is only carried out motion search to luminance component, the motion vector of output luminance component Y, the motion vector of the direct reference brightness component of the motion vector of chromatic component U and V Y, owing in addition the matching degree of colourity U and V component is calculated, in filter factor computational process, taken all factors into consideration Y, the matching degree of U and three components of V, carry out that motion compensated temporal filter can be eliminated because the distortion of distortion, particularly chromatic component that the matching degree of Y component and reference block and colourity U and V component and reference block matching degree do not occur simultaneously so use scheme provided by the invention in filtering.
Through the above description of the embodiments, those skilled in the art can be well understood to the present invention and can realize by hardware, also can realize by the mode that software adds necessary general hardware platform.Based on such understanding, technical scheme of the present invention can embody with the form of software product, it (can be CD-ROM that this software product can be stored in a non-volatile memory medium, USB flash disk, portable hard drive etc.) in, comprise some instructions with so that computer equipment (can be personal computer, server, the perhaps network equipment etc.) carry out the described method of each embodiment of the present invention.
More than disclosed only be several specific embodiment of the present invention, still, the present invention is not limited thereto, any those skilled in the art can think variation all should fall into protection scope of the present invention.
Claims (13)
1, a kind of method of filtering is characterized in that, comprising:
According to the combination of the matching degree of the matching degree of luminance component and chromatic component, adjust the initial filter coefficients of described luminance component and/or chromatic component;
Obtain total filter factor by the filter factor that passes through the described luminance component of adjusting and the filter factor of chromatic component.
2, the method for claim 1 is characterized in that, described according to described luminance component matching degree and the combination of the matching degree of described chromatic component, adjust before the initial filter coefficients of described luminance component and/or chromatic component, also comprise:
Luminance component in the digital video signal is separated with chromatic component;
In movement compensation process, luminance component is carried out motion search, obtain the match block that is complementary with described luminance component;
According to the luminance component matching degree of filtering image piece in described match block and the present frame, obtain initial filter coefficients shared weight in total filter factor of the motion vector and the described luminance component of described luminance component.
3, method as claimed in claim 2 is characterized in that, after the described motion vector that obtains described luminance component, also comprises:
Obtain the motion vector of chromatic component according to the motion vector of described luminance component;
Obtain the matching degree of described chromatic component by the motion vector of described chromatic component;
According to the matching degree of described chromatic component, obtain initial filter coefficients shared weight in total filter factor of described chromatic component.
4, method as claimed in claim 3 is characterized in that, described motion vector according to luminance component obtains the motion vector of chromatic component, comprising:
Form according to digital video signal is adjusted the motion vector of luminance component, obtains the motion vector that is complementary with described chromatic component.
5, the method for claim 1 is characterized in that, described according to described luminance component matching degree and the combination of the matching degree of described chromatic component, adjust the initial filter coefficients of described luminance component and/or chromatic component, comprising:
When the matching degree of a luminance component and two chromatic components all is to mate fully or when not matching fully, the initial filter coefficients of three components does not need to adjust.
6, the method for claim 1 is characterized in that, described according to described luminance component matching degree and the combination of the matching degree of described chromatic component, adjust the initial filter coefficients of described luminance component and/or chromatic component, comprising:
Have the matching degree of one-component to be in complete unmatched state in the matching degree of a luminance component and two chromatic components, the filter factor of three components all is adjusted into weak filter factor.
7, the method for claim 1 is characterized in that, described according to described luminance component matching degree and the combination of the matching degree of described chromatic component, adjust the initial filter coefficients of described luminance component and/or chromatic component, comprising:
Be to mate fully or do not match fully without any the matching degree of one-component in the matching degree of a luminance component and two chromatic components, the initial filter coefficients of the component that matching degree is bigger need be adjusted to the initial filter coefficients direction of the component of matching degree minimum.
8, the method for claim 1 is characterized in that, described filter factor by filter factor that passes through the described luminance component of adjusting and chromatic component obtains also to comprise after total filter factor:
Described luminance component and chromatic component are carried out time-domain filtering according to total filter factor;
Described luminance component behind the synthetic filtering and chromatic component and output.
9, a kind of device of filtering is characterized in that, comprising:
The filter factor adjusting module is used for the combination according to the matching degree of the matching degree of luminance component and chromatic component, adjusts the initial filter coefficients of described luminance component and/or chromatic component;
Total filter factor acquisition module is used for obtaining total filter factor by the filter factor that passes through the described luminance component of adjusting and the filter factor of chromatic component.
10, device as claimed in claim 9 is characterized in that, also comprises:
The digital video signal separation module is used for the luminance component of digital video signal is separated with chromatic component;
Brightness match block search module is used at movement compensation process luminance component being carried out motion search, obtains the match block that is complementary with described luminance component;
The brightness movement vector acquisition module is used for the matching degree according to described match block and present frame, obtains the motion vector of described luminance component;
Brightness initial filter coefficients weight acquisition module is used for the matching degree according to described match block and present frame, obtains initial filter coefficients shared weight in total filter factor of described luminance component.
11, device as claimed in claim 9, it is characterized in that, described filter factor adjusting module is when adjusting initial filter coefficients, when the matching degree of a luminance component and two chromatic components all is to mate fully or when not matching fully, the initial filter coefficients of three components does not need to adjust; Have the matching degree of one-component to be in complete unmatched state in the matching degree of a luminance component and two chromatic components, the filter factor of three components all is adjusted into weak filter factor; Be to mate fully or do not match fully without any the matching degree of one-component in the matching degree of a luminance component and two chromatic components, the initial filter coefficients of the component that matching degree is bigger need be adjusted to the initial filter coefficients direction of the component of matching degree minimum.
12, device as claimed in claim 9 is characterized in that, also comprises:
The chroma motion vector acquisition module is used for obtaining according to the motion vector of luminance component the motion vector of chromatic component;
Chromaticity match degree acquisition module is used for being obtained by described chroma motion vector the matching degree of described chromatic component;
Colourity initial filter coefficients weight acquisition module is used for obtaining described chromatic component initial filter coefficients in total shared weight of filter factor according to the matching degree of described chromatic component.
13, device as claimed in claim 9 is characterized in that, also comprises:
Brightness time-domain filtering module is used for described luminance component is carried out time-domain filtering according to total filter factor;
Colourity time-domain filtering module is used for described chromatic component is carried out time-domain filtering according to total filter factor;
Synthesis module is used for described luminance component and chromatic component and output behind the synthetic filtering.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200810146860 CN101662679B (en) | 2008-08-26 | 2008-08-26 | Method and device for filtering waves |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200810146860 CN101662679B (en) | 2008-08-26 | 2008-08-26 | Method and device for filtering waves |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101662679A true CN101662679A (en) | 2010-03-03 |
| CN101662679B CN101662679B (en) | 2011-12-21 |
Family
ID=41790400
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200810146860 Expired - Fee Related CN101662679B (en) | 2008-08-26 | 2008-08-26 | Method and device for filtering waves |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101662679B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107734211A (en) * | 2016-08-12 | 2018-02-23 | 瑞昱半导体股份有限公司 | Images filter method and its images filter device |
| CN113613005A (en) * | 2021-07-30 | 2021-11-05 | 百果园技术(新加坡)有限公司 | Video denoising method and device based on time domain filtering |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100367831B1 (en) * | 1998-06-30 | 2003-01-10 | 주식회사 팬택앤큐리텔 | Filtering method and apparatus of chrominance signal of an image |
| JP2004514351A (en) * | 2000-11-17 | 2004-05-13 | コーニンクレッカ フィリップス エレクトロニクス エヌ ヴィ | Video coding method using block matching processing |
| CN1181691C (en) * | 2003-01-24 | 2004-12-22 | 杭州国芯科技有限公司 | Vidio motion estimation method |
| CN101179726B (en) * | 2007-12-12 | 2010-09-29 | 北京中星微电子有限公司 | Motion estimating method and device |
-
2008
- 2008-08-26 CN CN 200810146860 patent/CN101662679B/en not_active Expired - Fee Related
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107734211A (en) * | 2016-08-12 | 2018-02-23 | 瑞昱半导体股份有限公司 | Images filter method and its images filter device |
| CN107734211B (en) * | 2016-08-12 | 2020-03-17 | 瑞昱半导体股份有限公司 | Image filtering method and image filtering device thereof |
| CN113613005A (en) * | 2021-07-30 | 2021-11-05 | 百果园技术(新加坡)有限公司 | Video denoising method and device based on time domain filtering |
| CN113613005B (en) * | 2021-07-30 | 2024-04-19 | 百果园技术(新加坡)有限公司 | Video denoising method and device based on time domain filtering |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101662679B (en) | 2011-12-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102231264B (en) | Dynamic contrast enhancement device and method | |
| CN102769722B (en) | Time-space domain hybrid video noise reduction device and method | |
| CN1992789B (en) | Motion estimator and motion method | |
| CN101841706B (en) | Method and device for improving video quality | |
| CN100366053C (en) | Method for compensating bad dots on digital images | |
| CN106504281B (en) | Image quality enhancing and filtering method applied to cmos image sensor | |
| CN107241607B (en) | Visual perception coding method based on multi-domain JND model | |
| US20070286287A1 (en) | Method and apparatus for improving quality of composite video signal and method and apparatus for removing artifact of composite video signal | |
| CN104363449B (en) | Image prediction method and relevant apparatus | |
| CN101924874A (en) | Matching block-grading realtime electronic image stabilizing method | |
| US20120294515A1 (en) | Image processing apparatus and image processing method, learning apparatus and learning method, program, and recording medium | |
| CN105763881A (en) | Video coding image filtering method based on image nonlocal structural similarity characteristic | |
| US20100111409A1 (en) | Image processing device, program recording medium, and image acquisition apparatus | |
| CN105635743A (en) | Minimum noticeable distortion method and system based on saliency detection and total variation | |
| EP4304172A1 (en) | Deblocking filter method and apparatus | |
| CN108833925B (en) | Inter-frame prediction method based on deep neural network | |
| CN101662679B (en) | Method and device for filtering waves | |
| US9041836B1 (en) | Digital picture noise reduction by combining high-noise and low-noise processed pictures | |
| US20120288209A1 (en) | Image processing apparatus and image processing method | |
| CN104486607B (en) | A kind of method and device of image chroma noise reduction | |
| JPH10191326A (en) | Image processor and method therefor | |
| US11202045B2 (en) | Image processing apparatus, imaging apparatus, image processing method, and program | |
| US8218896B2 (en) | Image display apparatus and method for correction chroma wrinkle | |
| CN107248147A (en) | The method that bad point is eliminated in RAW RGB images | |
| CN114827616B (en) | Compressed video quality enhancement method based on space-time information balance |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20111221 Termination date: 20130826 |