CN101964107A - Method for removing noise through pixel blocks - Google Patents

Abstract

The invention discloses a method for removing noise, which comprises the following steps of: dividing an image into pixel blocks with block areas and determining an adjacent block corresponding to each pixel block; defining and calculating difference between each adjacent pixel block and a central pixel block; further performing weighted value function processing on the difference between each adjacent pixel block and the central pixel block to acquire the weighted value of each adjacent pixel block and the central pixel block; and performing noise removal operation by using each weighted value of adjacent pixel blocks and the central pixel block and the pixel value of each pixel point of the central pixel block to acquire the pixel value of a central pixel point of the central pixel block. The method can keep the details of the image, well remove the noise and achieve good compromising effect.

Description

A kind of method of utilizing block of pixels to remove noise

Technical field

The present invention relates to a kind of method of removing noise, especially a kind of method of utilizing block of pixels to remove noise.

Background technology

In Flame Image Process, noise is a kind of common phenomena.Noise is big, influences picture quality.Therefore remove noise and be the important step in the Flame Image Process.

Prior art is a process object with a pixel, and also smoothly removes noise with the correlativity between the pixel.Because some pixel itself also is vulnerable to interference of noise, handles image with its correlativity and also can take noise in the further noise processed to.

Facts have proved that by multiple the place to go noise of prior art is for the details protection weak effect of image, and it is not good to remove anti noise under the prerequisite of disturbing existence.Such removal noise method is keeping image detail and is removing the effect that is difficult to reach a kind of compromise in the noise, so there is significant limitation in prior art.

Summary of the invention

In view of this, need a kind ofly can keep the removal noise technique that image detail can better be removed noise again and reach better compromise effect.The invention provides a kind of method of utilizing block of pixels to remove noise.

According to technical scheme of the present invention, provide a kind of noise remove method.This method comprises the steps:

Partitioned image is the block of pixels of boxed area, and the adjacent block of determining each block of pixels correspondence;

Each the adjacent block of pixels and the diversity factor of center pixel block are defined, and calculate;

Diversity factor to each adjacent block of pixels and center pixel block further is weighted the value function processing, obtains each the adjacent block of pixels and the weighted value of center pixel block;

Utilize the pixel value of each pixel of each weighted value of adjacent pixels piece and center pixel block and center pixel block, the computing of removing noise obtains the central pixel point pixel value of described center pixel block.

The present invention has overcome existing correlativity denoising method based on a pixel, to put pixel-expansion to block of pixels, and define a kind of diversity factor of new block of pixels, its pixel interference of noise of originally experiencing is weakened greatly, denoising effect also obtains bigger improvement, and in the end image detail has been carried out some corrections.Can better remove again in the balance of noise and reach compromise effect preferably keeping image detail.

Description of drawings

Below with reference to accompanying drawings specific embodiments of the present invention is described in detail, wherein:

Fig. 1 is a block of pixels synoptic diagram of the present invention;

Fig. 2 is a block of pixels diversity factor synoptic diagram of the present invention;

Fig. 3 is that the present invention removes block of pixels neighborhood synoptic diagram in the noise processed; And

Fig. 4 is that specific embodiment piece of the present invention is removed the noise processed process flow diagram.

Embodiment

In order to overcome the limitation of prior art, the invention provides a kind of noise remove method and device thereof based on the correlativity removal noise method of a pixel.Next will specify this method and device thereof.

Fig. 1 illustrates the synoptic diagram of block of pixels of the present invention.As shown in Figure 1, adopting 3*3 is example pixels illustrated piece, and nine pixels are formed a block of pixels altogether, that is:

str_pixel＝{pixel(a，b)，pixel(a，b-1)，pixel(a，b+1)，pixel(a-1，b)，pixel(a+1，b)

pixel(a-1，b-1)，pixel(a+1，b+1)，pixel(a-1，b+1)，pixel(a+1，b-1)}

The central pixel point of this block of pixels be pixel (a, b).

Being treated to of removal noise of the prior art: establishing a given image is h, and image is str_h after the denoising, and the denoising formula can be expressed as so:

$\mathrm{str}\_h\left(i\right)=\underset{j\∈I}{\mathrm{\Σ}}w(i,j)h\left(j\right)---\left(2\right)$

Wherein I represents the set of pixels relevant with the i pixel, and w () is the weighted value of respective pixel, and existing algorithm all is to be defined as a pixel with the j pixel.

In the present invention, the principle of setting block of pixels is:

The pixel region of a m*n of definition is a block of pixels because be that to have local image structure information be image detail in zone, then block of pixels as a some pixel, the computing that substitution formula (2) is removed noise.

Fig. 2 illustrates block of pixels diversity factor synoptic diagram of the present invention.As shown in Figure 2, two block of pixels str_pixel (a, b) and str_pixel (a, b-1), the span of two pixels is as shown in the figure 9 points respectively, the diversity factor of these two block of pixels is calculated as follows:

$\mathrm{Diff}(\mathrm{str}\_\mathrm{pixel}(a,b),\mathrm{str}\_\mathrm{pixel}(a,b-1))=\frac{\sqrt{\left[\underset{a=-1}{\overset{1}{\mathrm{\Σ}}}\underset{b=-1}{\overset{1}{\mathrm{\Σ}}}{\left(\mathrm{pixel}(a,b)\right)}^2\underset{a=-1}{\overset{1}{\mathrm{\Σ}}}\underset{b=-2}{\overset{0}{\mathrm{\Σ}}}{\left(\mathrm{pixel}(a,b)\right)}^2\right]}}{\underset{a=-1}{\overset{1}{\mathrm{\Σ}}}\underset{b=-1}{\overset{1}{\mathrm{\Σ}}}\left[\mathrm{pixel}(a,b)\right]\left[\mathrm{pixel}(a,b-1)\right]}---\left(3\right)$

For two block of pixels str_pixel_i arbitrarily, the diversity factor of str_pixel_j is calculated, and available following general formula is represented:

$\mathrm{Diff}(\mathrm{str}\_\mathrm{pixel}\_i,\mathrm{str}\_\mathrm{pixel}\_j)=\frac{\sqrt{\left[{\mathrm{\Σ}}_a{\mathrm{\Σ}}_b{(\mathrm{pixel}\_i(a,b))}^2{\mathrm{\Σ}}_a{\mathrm{\Σ}}_b{(\mathrm{pixel}\_j(a,b))}^2\right]}}{{\mathrm{\Σ}}_a{\mathrm{\Σ}}_b[\mathrm{pixel}\_i(a,b)][\mathrm{pixel}\_j(a,b)]}---\left(4\right)$

Fig. 3 illustrates the present invention and removes block of pixels neighborhood synoptic diagram in the noise processed.As shown in Figure 3, Fig. 3 (a)-Fig. 3 (i) is the square formation that 3*3 block of pixels constitutes, and Fig. 3 (e) is a center pixel block, and its neighborhood is to comprise center pixel block in interior 9 block of pixels as shown in FIG..

As can be seen from the figure, block of pixels with and the adjacent pixels piece the overlapping phenomenon of pixel is all arranged when dividing.This division is just removed association relative in the noise operation and is divided, and does not relate to the structure that changes image and the numerical value of pixel.

The adjacent pixel blocks of block of pixels of the present invention is not less than 3*3 block of pixels.

Fig. 4 illustrates specific embodiment of the present invention and removes the noise processed process flow diagram.As shown in Figure 4, step 400 beginning, in step 402, input picture, the definition block of pixels is carried out the division of block of pixels and the adjacent pixel blocks of determining each block of pixels correspondence, as above shown in the formula (1).

In step 404, image boundary is handled, the boundary treatment in this step be in image boundary when dividing the block of pixels of pixel, lack the pixel that makes up block of pixels, need the pixel that lack be made up by mirror image processing.

Repeat to handle for image boundary, promptly a kind of mirror image processing, its corresponding relation is as follows:

v(i，n+1)＝v(i，n)

v(i，0)＝v(i，1)

v(0，j)＝v(1，j) (5)

v(m+1，j)＝v(m，j)

In the following formula, M, N are respectively the height and the width of image, i, and the span of j is respectively [1, M] and [1, N].

Further, in step 406, determine the right block of pixels diversity factor of block of pixels that center pixel block is adjacent, this diversity factor is meant the difference of two block of pixels, as above shown in the formula (4).

Then, in step 408, the right weighted value function of block of pixels that center pixel block is adjacent is set.

It is as follows that weighted value function f () can be set:

$f(i,j)=\frac{1}{\sqrt{2\mathrm{\π}}\mathrm{\σ}}\exp (-\frac{\mathrm{Diff}{(\mathrm{str}\_\mathrm{pixel}\_i-\mathrm{str}\_\mathrm{pixel}\_j)}^2}{{\mathrm{\σ}}^2})---\left(6\right)$

Wherein, the σ noise variance, for this method, can be by being set to artificially reconcile parameter.The definition of diversity factor in the following formula is calculated according to (4) formula, and the weighted value function is to determine its contribution margin to center pixel according to the difference between adjacent pixel blocks and the center pixel block, and diversity factor is big more, contributes more for a short time, and diversity factor is more little, contributes big more.

For each pixel, the adjacent pixel blocks of its block of pixels is example with 3*3, and the weighted value function that the block of pixels that its center pixel block is adjacent is right has 9.

In step 410, the weighted value that the block of pixels that center pixel block is adjacent is right is carried out normalized, and its processing is as follows:

$f^{\′}(i,j)=\frac{f(i,j)}{\underset{i,j}{\mathrm{\Σ}}f(i,j)}---\left(7\right)$

Because the pixel value value of image within the specific limits, so weighted value need be carried out normalization and could do not changed the span of removing the pixel value behind the noise.

And then, enter in the step 412, remove the calculation process of noise, with the corresponding product of pixel value of each weighted value of adjacent pixels piece and center pixel block and each pixel of center pixel block and sue for peace, remove The noise, obtain the pixel value of block of pixels central pixel point.

For example in Fig. 3, pixel (a, the denoising result of b) locating can be expressed as:

str_w(a，b)＝f′ _aew(a-1，b-1)+f′ _bew(a-1，b)+f′ _cew(a-1，b+1)

f′ _dew(a，b-1)+f′ _eew(a，b)+f′ _few(a，b+1) (8)

f′ _gew(a+1，b-1)+f′ _hew(a+1，b)+f′ _iew(a+1，b+1)

In the formula, w (x, y) remarked pixel (x, the y) value at coordinate place, w ' expression weights, its subscript remarked pixel piece is right, particularly, w ' _AeRepresent Fig. 3 (a) and the right treated normalization weighted value of this a pair of block of pixels of 3 (e) as shown in Figure 3, other weights refer to that other block of pixels is right.

In step 414, the pixel information of output image enters then in the step 416 and finishes.

More than specific descriptions of the present invention are intended to illustrate the implementation of specific embodiments can not be interpreted as it is limitation of the present invention.Those of ordinary skills can make various variants on the basis of the embodiment that describes in detail under instruction of the present invention, these variants all should be included within the design of the present invention.The present invention's scope required for protection is only limited by described claims.

Claims (6)

1. method of utilizing block of pixels to remove noise may further comprise the steps:

Partitioned image is the block of pixels of boxed area, and the adjacent block of determining each block of pixels correspondence;

Each the adjacent block of pixels and the diversity factor of center pixel block are defined, and calculate;

Diversity factor to each adjacent block of pixels and center pixel block further is weighted the value function processing, obtains each the adjacent block of pixels and the weighted value of center pixel block;

Utilize the pixel value of each pixel of each weighted value of adjacent pixels piece and center pixel block and center pixel block, the computing of removing noise obtains the central pixel point pixel value of described center pixel block.

2. the method for removal noise as claimed in claim 1 is characterized in that, described each the adjacent block of pixels and the diversity factor of center pixel block adopt as giving a definition:

$\mathrm{Diff}(\mathrm{str}\_\mathrm{pixel}\_i,\mathrm{str}\_\mathrm{pixel}\_j)=\frac{\sqrt{\left[{\mathrm{\Σ}}_a{\mathrm{\Σ}}_b{(\mathrm{pixel}\_i(a,b))}^2{\mathrm{\Σ}}_a{\mathrm{\Σ}}_b{(\mathrm{pixel}\_j(a,b))}^2\right]}}{{\mathrm{\Σ}}_a{\mathrm{\Σ}}_b[\mathrm{pixel}\_i(a,b)][\mathrm{pixel}\_j(a,b)]}.$

3. noise remove method as claimed in claim 1 is characterized in that, described weighted value function is:

$f(i,j)=\frac{1}{\sqrt{2\mathrm{\π}}\mathrm{\σ}}\exp (-\frac{\mathrm{Diff}{(\mathrm{str}\_\mathrm{pixel}\_i-\mathrm{str}\_\mathrm{pixel}\_j)}^2}{{\mathrm{\σ}}^2}).$

4. noise remove method as claimed in claim 1 is characterized in that:

Described each adjacent block of pixels and center pixel block are weighted after value function handles, also comprise the weighted value that obtains is carried out normalized.

5. as each described noise remove method of claim 1 to 4, it is characterized in that:

Described block of pixels is the pixel square formation that is not less than 3*3.

6. as each described noise remove method of claim 1 to 4, it is characterized in that:

The number of block of pixels is for being not less than 3*3 in the adjacent pixel blocks of described block of pixels.

Images