CN102547068A - Improved bilinear interpolation video scaling method - Google Patents

Abstract

The invention discloses an improved bilinear interpolation video scaling method which comprises the following steps of: firstly, receiving a video signal to be processed; secondly, judging whether the video signal is required to be subjected to amplifying processing or scaling processing, if the video signal is subjected to the amplifying processing, carrying out sharpening filter processing on the received video signal, then carrying out bilinear interpolation processing on the video signal subjected to sharpening filter processing and finally outputting the video signal subjected to bilinear interpolation processing, and otherwise, carrying out bilinear interpolation processing on the received video signal first, then carrying out sharpening filter processing on the video signal subjected to bilinear interpolation processing and finally outputting the video signal subjected to sharpening filter processing. According to the improved bilinear interpolation video scaling method disclosed by the invention, the mode of carrying out sharpening filter processing by adopting a combined filter, a log filter or a laplace filter is preferably considered; a preposed sharpening filter is used for carrying out amplifying processing on an video image and a postposed sharpening filter is used for carrying out minifying processing on the video image, so that the quality of bilinear interpolation video scaling can be improved, edge blur and sawtooth in the scaling processing process of the video image can be effectively overcome and the video image with better image quality can be obtained.

Description

Improved bilinear interpolation video scaling method

Technical field

The invention belongs to the digital video image process field, specifically, relate to a kind of edge blurry and improved bilinear interpolation video scaling method of crenellated phenomena that can effectively overcome in the video image zooming processing procedure.

Background technology

Image zoom is widely used in digital picture and the video processing applications; Comprise fields such as consumer electronics product and medical image; Along with various consumer electronics products extensively adopt flat-panel display device; Like mobile phone, video player, panel computer and LCD TV etc., on the flat-panel display device of various different resolutions, show with a kind of video source of resolution, must adopt the image zoom technology.

Present image zoom algorithm mainly divides three types: 1) traditional interpolation algorithm; 2) based on the interpolation algorithm at edge; 3) based on the interpolation algorithm of estimation.The two kinds of algorithms in back seldom adopt in consumer electronics product because operand is big, and traditional interpolation algorithm is widely used in the consumer electronics product because operand is less, arest neighbors interpolation, bilinear interpolation and the Tri linear interpolation of mainly containing commonly used.Tradition interpolation image convergent-divergent algorithm all is that to be based upon band-limited signal be perfectly to rebuild through ideal low-pass filter; Therefore the low pass filter of these algorithm design all is to level off to the frequency spectrum of perfect low pass filtering; Yet; Picture signal is with limit, so traditional interpolation algorithm can not recover the HFS of image, finally causes edge of image fuzzy or have a crenellated phenomena.

Summary of the invention

To above deficiency, the invention provides a kind of edge blurry and improved bilinear interpolation video scaling method of crenellated phenomena that can effectively overcome in the video image zooming processing procedure, it at first receives pending vision signal; Then judge it is that this vision signal is carried out processing and amplifying, still carry out convergent-divergent and handle; If processing and amplifying; Earlier the vision signal that receives is carried out the sharpening Filtering Processing, and then the vision signal after the sharpening Filtering Processing is carried out bilinear interpolation handle the output video image that disposes at last; If convergent-divergent is handled; Earlier the vision signal that receives is carried out bilinear interpolation and handle, then bilinear interpolation processed video signal is carried out the sharpening Filtering Processing, the output video image that disposes at last.

The present invention adopts the sharpening filter to carry out the sharpening Filtering Processing.

Said sharpening filter is an associated filters, and its filter factor is:

$\mathrm{Kernel}=\left[\begin{array}{ccccc}-1& -2& -3& -2& -1\\ -2& -2-C+S& -4-C+S& -2-C+S& -2\\ -3& -4-C+S& -8+\mathrm{SC}& -4-C+S& -3\\ -2& -2-C+S& -4-C+S& -2-C+S& -2\\ -1& -2& -3& -2& -1\end{array}\right]/(C+8)(S+8)$

Wherein, C and S are parameters, and their value is adjusted according to feature of image.

Said sharpening filter is the logarithmic filtering device, and its filter factor is:

${\mathrm{Kernel}}_{\mathrm{LOC}}=\left[\begin{array}{ccc}-1& -2& -1\\ -2& B& -2\\ -1& -2& -1\end{array}\right]$

Wherein, B is the sharpening parameter, adjusts according to characteristics of image, must be greater than-12.

Said sharpening filter is a Laplace filter, and its filter factor is:

${\mathrm{Kernel}}_{\mathrm{LAP}}=\left[\begin{array}{ccc}0& -1& 0\\ -1& C& -1\\ 0& -1& 0\end{array}\right]$

Wherein, C is the sharpening parameter, adjusts according to characteristics of image, must be greater than-4.

Beneficial effect of the present invention: the present invention adopts preposition sharpening filter for the processing and amplifying of video image; And being dwindled processing, video image adopts rearmounted sharpening filter; Can promote the quality of bilinear interpolation video scaling significantly; Can effectively overcome edge blurry and crenellated phenomena in the video image zooming processing procedure, obtain the video image of better image quality.

Description of drawings

Fig. 1 is the improved bilinear interpolation video scaling of a present invention method flow diagram;

Fig. 2 is a bilinear interpolation algorithm principle sketch map.

Embodiment

Below in conjunction with accompanying drawing the present invention is further set forth.

As shown in Figure 1, improved bilinear interpolation video scaling method of the present invention comprises:

1) receives pending vision signal.

2) judge it is that this vision signal is carried out processing and amplifying, still carry out convergent-divergent and handle, if processing and amplifying, execution in step 3), deny the person, execution in step 4).

3) vision signal that receives is carried out the sharpening Filtering Processing, and then the vision signal after the sharpening Filtering Processing is carried out bilinear interpolation handle, export final video image at last, termination routine.

4) vision signal that receives is carried out bilinear interpolation and handle, then bilinear interpolation processed video signal is carried out the sharpening Filtering Processing, export final video image at last, termination routine.

Be that the present invention adopts the bilinear interpolation algorithm that video image is carried out before the convergent-divergent, at first judge it is that video image is amplified or dwindles; Then,, then earlier raw video image is carried out sharpening filtering, carry out bilinear interpolation again if raw video image is amplified; If raw video image is dwindled, carry out bilinear interpolation earlier, carry out sharpening filtering again.

Because bilinear interpolation has the character of low pass filter; Make high fdrequency component impaired, thus image outline is thickened, therefore; The present invention adopts the sharpening filter to improve the quality of image zoom; The kind of sharpening filter is a lot, and the present invention is through the test of multitude of video image zoom, and the present invention pays the utmost attention to and adopts the apparent in view sharpening filter of following three kinds of effects: associating sharpening filter, logarithm sharpening filter, Laplce's sharpening filter.

Filter factor is also claimed operator or convolution template, and the concrete coefficient of associated filters is:

$\mathrm{Kernel}=\left[\begin{array}{ccccc}-1& -2& -3& -2& -1\\ -2& -2-C+S& -4-C+S& -2-C+S& -2\\ -3& -4-C+S& -8+\mathrm{SC}& -4-C+S& -3\\ -2& -2-C+S& -4-C+S& -2-C+S& -2\\ -1& -2& -3& -2& -1\end{array}\right]/(C+8)(S+8)$

Wherein, C and S are parameters, can adjust according to feature of image.

Logarithm sharpening filter filtering coefficient is:

${\mathrm{Kernel}}_{\mathrm{LOC}}=\left[\begin{array}{ccc}-1& -2& -1\\ -2& B& -2\\ -1& -2& -1\end{array}\right]$

In the following formula, B is the sharpening parameter, can adjust according to characteristics of image, must be greater than-12.

The Laplace filter filter factor is:

${\mathrm{Kernel}}_{\mathrm{LAP}}=\left[\begin{array}{ccc}0& -1& 0\\ -1& C& -1\\ 0& -1& 0\end{array}\right]$

In the following formula, C is the sharpening parameter, can adjust according to characteristics of image, must be greater than-4.

Bilinear interpolation algorithm (Bilinear Interpolation) is modal linear interpolation; Because can linear interpolation be decomposed into vertical and horizontal both direction carries out respectively; So be called bilinear interpolation, its video image zooming quality will be higher than neighbor interpolation algorithm.It obtains gray values of pixel points to be inserted according to the gray value of the related pixel point of 4 neighborhoods around the pixel to be inserted through the two-dimensional linear weighted average calculation, and shown in accompanying drawing 2, the value of the pixel of asking is:

P(xx，yy)＝(1-Δy)·[(1-Δx)·f(x，y)+Δx·f(x+1，y)]+Δy·[(1-Δx)·f(x，y+1)+Δx·f(x+1，y+1)]

Xx wherein, yy is the location of pixels of output image.F (x, y), f (x+1, y), f (x, y+1) and f (x+1 y+1) is the pixel value in neighbours territory in the original image.If separate horizontal scaling and vertically scale, the convergent-divergent row convergent-divergent again of going ahead of the rest, so above-mentioned formula can be write:

I(xx，y)＝(1-Δx)·f(x，y)+Δx·f(x+1，y)；

I(xx，y+1)＝(1-Δx)·f(x，y+1)+Δx·f(x+1，y+1)；

P(xx，yy)＝(1-Δy)·I(xx，y)+Δy·I(xx，y+1)；

Wherein, I () has represented the interpolation calculation intermediate point of two grey up and down.

The above is merely preferred embodiments of the present invention; The present invention is not limited to above-mentioned execution mode; In implementation process, possibly there is local small structural modification; If various changes of the present invention or modification are not broken away from the spirit and scope of the present invention, and belong within claim of the present invention and the equivalent technologies scope, then the present invention also is intended to comprise these changes and modification.

Claims (5)

1. an improved bilinear interpolation video scaling method is characterized in that it at first receives pending vision signal; Then judge it is that this vision signal is carried out processing and amplifying, still carry out convergent-divergent and handle; If processing and amplifying; Earlier the vision signal that receives is carried out the sharpening Filtering Processing, and then the vision signal after the sharpening Filtering Processing is carried out bilinear interpolation handle the output video image that disposes at last; If convergent-divergent is handled; Earlier the vision signal that receives is carried out bilinear interpolation and handle, then bilinear interpolation processed video signal is carried out the sharpening Filtering Processing, the output video image that disposes at last.

2. improved bilinear interpolation video scaling method according to claim 1, it adopts the sharpening filter to carry out the sharpening Filtering Processing.

3. improved bilinear interpolation video scaling method according to claim 2, said sharpening filter is an associated filters, its filter factor is:

$\mathrm{Kernel}=\left[\begin{array}{ccccc}-1& -2& -3& -2& -1\\ -2& -2-C+S& -4-C+S& -2-C+S& -2\\ -3& -4-C+S& -8+\mathrm{SC}& -4-C+S& -3\\ -2& -2-C+S& -4-C+S& -2-C+S& -2\\ -1& -2& -3& -2& -1\end{array}\right]/(C+8)(S+8)$

Wherein, C and S are parameters, and their value is adjusted according to feature of image.

4. improved bilinear interpolation video scaling method according to claim 2, said sharpening filter is the logarithmic filtering device, its filter factor is:

${\mathrm{Kernel}}_{\mathrm{LOG}}=\left[\begin{array}{ccc}-1& -2& -1\\ -2& B& -2\\ -1& -2& -1\end{array}\right]$

Wherein, B is the sharpening parameter, adjusts according to characteristics of image, must be greater than-12.

5. improved bilinear interpolation video scaling method according to claim 2, said sharpening filter is a Laplace filter, its filter factor is:

${\mathrm{Kernel}}_{\mathrm{LAP}}=\left[\begin{array}{ccc}0& -1& 0\\ -1& C& -1\\ 0& -1& 0\end{array}\right]$

Wherein, C is the sharpening parameter, adjusts according to characteristics of image, must be greater than-4.

Images