CN1874450A - Method and apparatus for adjusting brightness of image - Google Patents

Abstract

A method and apparatus for adjusting brightness of an image are provided, in which number distribution information of pixel brightness is detected, and the brightness of a specific area of the image is adjusted using characteristic values according to a distribution characteristic of the detected number distribution information. Accordingly, a decrease in an image recognition characteristic of a human being is compensated for in a specific area, such as, low-brightness area, of an image.

Description

Be used to regulate the method and apparatus of image brightness

The cross reference of related application

It is the priority of 2005112311 Russ P application at the application number of Russ P office application that the application requires on April 25th, 2005, and on March 29th, 2006 be the priority of the korean patent application of 10-2006-0028423 at the application number of Korea S Department of Intellectual Property application, whole disclosures of above-mentioned application are included in this for your guidance.

Technical field

The present invention relates to image processing, more specifically, relate to a kind of method and apparatus of brightness of the specific region that is used to regulate image.

Background technology

Recently, image processing apparatus, for example digital camera and camcorder (camcorder) make the user be easy to obtain high-resolution image.Usually, these images are loaded into personal computer (PC) and are edited, and are stored in subsequently in the memory space of PC or by the printing equipment of for example printer or multi-function peripheral device (MFP) and print.

But under the situation of such image, in a specific region, the height of brightness and low being difficult to are distinguished.Especially, because human vision property, a low brightness area (black region) may be difficult to distinguish.Therefore, in the low brightness area of image, bright and secretly can not clearly distinguish.If increase the brightness of image fully, then the brightness of the high-brightness region of image also can increase.Therefore, high-brightness region will be shown as the white of can't be as image and being recognized.

Therefore, need a kind of the improving one's methods and installing of brightness of the specific region that be used to regulate image.

Summary of the invention

An aspect of one exemplary embodiment of the present invention is will address the above problem at least and/or shortcoming, and following advantage is provided at least.Correspondingly, one exemplary embodiment of the present invention aspect provides the method and apparatus that a kind of brightness that is used to regulate image comes the compensator that the image identifying characteristic of a specific region is descended.

An aspect according to one exemplary embodiment of the present invention provides a kind of method of regulating image brightness, and this method comprises: detect quantity (number) distributed intelligence of pixel brightness; Utilize characteristic value to regulate the brightness of the specific region of image with distribution character according to detected distributed number information.

According to another aspect of one exemplary embodiment of the present invention, a kind of device that is used to regulate image brightness is provided, this device comprises: a distributed intelligence detector that detects the distributed number information of pixel brightness; Utilize characteristic value to regulate the brightness regulator of brightness of the specific region of image with a distribution character according to detected distributed number information.

According in conjunction with the accompanying drawings to disclosing the following detailed description of one exemplary embodiment of the present invention, to one skilled in the art, other purpose of the present invention, advantage and prominent features will become clearly.

Description of drawings

According to following description in conjunction with the accompanying drawings, make one exemplary embodiment more of the present invention above-mentioned and other purpose, feature and advantage become more clear, wherein:

Accompanying drawing 1 is the flow chart according to the method for the adjusting image brightness of an one exemplary embodiment of the present invention;

Accompanying drawing 2 is detail flowcharts that the operation 14 in the accompanying drawing 1 is shown;

Accompanying drawing 3A and 3B illustrate the histogram (histogram) of the distributed number information of an image and this image;

Accompanying drawing 4 is detail flowcharts that the operation 16 in the accompanying drawing 1 is shown;

Accompanying drawing 5 is schematic diagrames that are shown the distributed number information of histogram, and it is divided into first to the 3rd cut zone based on brightness;

Accompanying drawing 6 is detail flowcharts that the operation 52 of accompanying drawing 4 is shown;

Accompanying drawing 7A-D illustrates the image that obtains by the operation of carrying out accompanying drawing 4;

Accompanying drawing 8 is schematic diagrames that one exemplary embodiment is used for the image brightness of comparison specific region according to the present invention;

Accompanying drawing 9 is block diagrams of regulating the device of image brightness according to one exemplary embodiment of the present invention;

Accompanying drawing 10 is block diagrams of the contrast adjustment device of accompanying drawing 9; With

Accompanying drawing 11 is block diagrams of the brightness regulator of accompanying drawing 9.

Spread all over each accompanying drawing, should understand identical Reference numeral and refer to components identical, feature and structure.

Specific embodiment

The for example detailed structure of qualification and the content of element are provided in following description, help the complete understanding embodiments of the invention.Therefore, it will be appreciated by those skilled in the art that under the situation that does not depart from the scope of the present invention and conceive, can carry out various changes and improvements wherein said embodiment.In addition, for the sake of simplicity and brief description of omitting to known function and structure.

Accompanying drawing 1 is an one exemplary embodiment according to the present invention to be shown regulate the flow chart of the method for image brightness.

With reference to the accompanying drawings 1, in operation 10, the distributed number information of the pixel brightness of detected image.Distributed number information is the information about the quantity of the pixel with same brightness, and it comes the pixel of classified image to obtain by the brightness according to each pixel.Therefore, distributed number information is suitable for being illustrated by a Nogata.

In operation 12, determine whether to carry out the contrast adjustment of image.If do not need the contrast adjustment of image, this processing proceeds to operation 16, and it utilizes characteristic value to regulate the brightness of the specific region of image.The contrast adjustment of image in operation 14, is regulated the contrast of image if desired.

Accompanying drawing 2 is detail flowcharts that the operation 14 of accompanying drawing 1 is shown.

With reference to the accompanying drawings 2, in operation 30, from image, detect first reference brightness value of an indication low-light level and the second reference brightness value of an indication high brightness.

When one in the brightness value brightness value corresponding to minimum pixel number during less than a first threshold brightness value, wherein pixel number is greater than first threshold quantity, then is taken as the first reference brightness value less than the brightness value of first threshold brightness value and detects.When the low brightness area from image obtains the first reference brightness value, first threshold quantity is a predetermined quantity that is used as the comparison reference, and when the low brightness area from image obtained the first reference brightness value, the first threshold brightness value was one and is used as another relatively predetermined brightness value of reference.Utilize formula 1 to obtain the first reference brightness value.

Low=min (i): H (i)＞N ₁And i≤T ₁(1)

In formula 1, low represents the first reference brightness value, and H (i) expression is corresponding to the pixel number of brightness value i, N ₁Expression first threshold quantity, and T ₁Expression first threshold brightness value.

In addition, when one in the brightness value brightness value corresponding to maximum pixel quantity during greater than one second threshold luminance value, wherein pixel number is greater than one second number of thresholds, is taken as the second reference brightness value greater than the brightness value of the second threshold luminance value and detects.When obtaining the second reference brightness value in the high-brightness region at image, second number of thresholds is a predetermined quantity that is used as the comparison reference, and when the high-brightness region at image obtained the second reference brightness value, the second threshold luminance value was one and is used as another relatively predetermined brightness value of reference.Utilize formula 2 to obtain the second reference brightness value.

High=max (i): H (i)＞N ₂And i≤T ₂(2)

In formula, high represents the second reference brightness value, and H (i) expression is corresponding to the pixel number of brightness value i, N ₂Represent second number of thresholds, and T ₂Represent the second threshold luminance value.

Accompanying drawing 3A and 3B illustrate the histogram of the distributed number information of an image and this image.Accompanying drawing 3A illustrates this image, and accompanying drawing 3B is a histogram with the distributed number information representation of the image among the accompanying drawing 3A.If suppose T ₁Be a first threshold brightness value, and N ₁Be a first threshold quantity, less than T ₁Brightness value in a brightness value corresponding to minimum pixel number be taken as the first reference brightness value and detect, wherein pixel number is greater than N ₁In addition, if hypothesis T ₂Be the second threshold luminance value, and N ₂Be second number of thresholds, greater than T ₂Brightness value in a brightness value corresponding to maximum pixel quantity be taken as the second reference brightness value and detect, wherein pixel number is greater than N ₂

In operation 32, utilize the detected first and second reference brightness values to regulate R, G and the B of image.

Can use formula 3 to regulate the R of image, G and B.

R′(r，c)＝(2 ⁿ-1)*(R(r，c)-low)/(high-low)

G′(r，c)＝(2 ⁿ-1)*(G(r，c)-low)/(high-low) (3)

B′(r，c)＝(2 ⁿ-1)*(B(r，c)-low)/(high-low)

Wherein, the coordinate of each pixel in r and the c presentation video, n are represented the constant of the quantity of an indication brightness degree (gradation), and low represents the first reference brightness value, and high represents the second reference brightness value.

By utilizing formula 3 to regulate new R ', G ' and B ' regulate the contrast of image.For example, but in one exemplary embodiment of the present invention the contrast of broadening image.

In operation 34, detect R, the distributed number information of G and B image through regulating.R, the distributed number information of G and the B image through regulating can detect as histogram information according to above-mentioned.

Return with reference to the accompanying drawings 1, in operation 16, utilize the brightness of regulating the image of specific region according to the characteristic value of the distribution character of the distributed number information that detects.Example according to the characteristic value of the distribution character of the distributed number information that detects will be described below.

Accompanying drawing 4 is detail flowcharts of the operation 16 of accompanying drawing 1.

With reference to the accompanying drawings 4, detected characteristics value in operation 50.

The distributed number information that will be expressed as histogram based on brightness is divided into a plurality of zones, and detects the characteristic value of each cut zone.Especially, a plurality of zones are split into first cut zone, second cut zone and the 3rd cut zone.But this Region Segmentation only is a kind of diagram, and histogram can be split into multiple zone.

Accompanying drawing 5 is schematic diagrames that are represented as the distributed number information of histogram, and wherein histogram is split into first to the 3rd cut zone based on brightness.As shown in the accompanying drawing 5, histogram is split into the first cut zone A based on brightness ₁, the second cut zone A ₂And the 3rd cut zone A ₃The first cut zone A ₄Be low brightness area, the second cut zone A ₂Be an intermediate light zone, and the 3rd cut zone A ₃It is a high-brightness region.The first cut zone A ₁Be divided into two zones, one first subregion A ₁₁With one second subregion A ₁₂

Formula 4 to 8 can be used for detecting the characteristic value of cut zone.

$\mathrm{gm}=\underset{i\∈[0,2^n-1]}{\max \left(H\left(i\right)\right)}---\left(4\right)$

In formula 4, H (i) expression is corresponding to the pixel number of brightness value i, and gm represents corresponding to the maximum pixel quantity in the pixel number of the brightness value of image, and n represents to indicate the constant of brightness degree quantity.For example, when n=8, gm is corresponding to the maximum pixel quantity in the pixel number of the brightness value of the image with 0 to 255 brightness degree.

${\mathrm{<figure-callout\; id="1"\; label="S"\; filenames="A20061009986700221.png"\; state="\{\{state\}\}">S</figure-callout>}}_1=\underset{\mathrm{first}}{\mathrm{\&Sigma;}}H\left(i\right)/(M*N),{\mathrm{<figure-callout\; id="2"\; label="S"\; filenames="A20061009986700251.png"\; state="\{\{state\}\}">S</figure-callout>}}_2=\underset{\sec \mathrm{ond}}{\mathrm{\&Sigma;}}H\left(i\right)/(M*N),{\mathrm{<figure-callout\; id="3"\; label="S"\; filenames="A20061009986700251.png"\; state="\{\{state\}\}">S</figure-callout>}}_3=\underset{\mathrm{third}}{\mathrm{\&Sigma;}}H\left(i\right)/(M*N)---\left(5\right)$

In formula 5, S ₁Expression is corresponding to the first cut zone A in the entire image ₁The pixel number ratio of brightness, S ₂Expression is corresponding to the second cut zone A in the entire image ₂The pixel number ratio of brightness, S ₃Expression is corresponding to the 3rd cut zone A in the entire image ₃The pixel number ratio of brightness, and the size of M*N presentation video (resolution).

$S_{11}=\underset{i=0}{\overset{\{(2^n-1)/6\}-1}{\mathrm{\&Sigma;}}}H\left(i\right)/(M*N),S_{12}=\underset{i=(2^n-1)/6}{\overset{(2^n-1)/3}{\mathrm{\&Sigma;}}}H\left(i\right)/(M*N),---\left(6\right)$

As the first cut zone A ₁Be split into the first subregion A ₁₁With the second subregion A ₁₂, in formula 6, S ₁₁Expression is corresponding to the first subregion A in the entire image ₁₁The pixel number ratio of brightness, and S ₁₂Expression is corresponding to the second subregion A in the entire image ₁₂The pixel number ratio of brightness.

${\mathrm{<figure-callout\; id="1"\; label="R"\; filenames="A20061009986700221.png"\; state="\{\{state\}\}">R</figure-callout>}}_1=\underset{\mathrm{first}}{\max \left(H\left(i\right)\right)/\mathrm{gm},{\mathrm{<figure-callout\; id="2"\; label="R"\; filenames="A20061009986700251.png"\; state="\{\{state\}\}">R</figure-callout>}}_2=}\underset{\sec \mathrm{<figure-callout\; id="3"\; label="ond\; R"\; filenames="A20061009986700251.png"\; state="\{\{state\}\}">ond</figure-callout>}}{\max \left(H\left(i\right)\right)/\mathrm{gm},}{R}_{}{}_3=\underset{\mathrm{third}}{\max \left(H\left(i\right)\right)/\mathrm{gm},---\left(7\right)}$

In formula 7, R ₁Be illustrated in corresponding to the first cut zone A ₁In the pixel number of brightness value in the ratio of maximum pixel quantity and gm, gm is corresponding to the maximum pixel quantity in the pixel number of the brightness value of image, R ₂Be illustrated in corresponding at the second cut zone A ₂In the pixel number of brightness value in the ratio of maximum pixel quantity and gm, and R ₃Be illustrated in corresponding at the 3rd cut zone A ₃In the pixel number of brightness value in the ratio of maximum pixel quantity and gm.

P ₁＝h/(2 ⁿ-1)，P ₂＝l/(2 ⁿ-1) (8)

In formula 8, h represents to have the first cut zone A ₁In the brightness value of maximum pixel quantity, P ₁Expression h and 2 ⁿThe ratio of grade, l represent that has the 3rd a cut zone A ₃In the brightness value of maximum pixel quantity, P ₂Expression l and 2 ⁿThe ratio of grade.

In operation 52, utilize detected characteristic value to detect a coefficient a who is used to regulate the brightness of specific region.

Accompanying drawing 6 is detail flowcharts that the operation 52 of accompanying drawing 4 is shown.

With reference to the accompanying drawings 6, in operation 70, determine characteristic value R ₃Whether greater than characteristic value R ₂With characteristic value R ₁

If determine characteristic value R ₃Greater than characteristic value R ₂With characteristic value R ₁, in operation 71, determine characteristic value P ₂Whether greater than 0.98* (2 ⁿ-1).

If determine characteristic value P ₂Greater than 0.98* (2 ⁿ-1), in operation 72, determines characteristic value P ₁Whether less than 0.1* (2 ⁿ-1), S ₁/ S ₂Whether greater than 1.3 and S ₁₁Whether greater than S ₁₂

If determine characteristic value P ₁Less than 0.1* (2 ⁿ-1), S ₁/ S ₂Greater than 1.3, and S ₁₁Greater than S ₁₂, in operation 73, coefficient a is detected as 0.7.

If do not determine characteristic value P ₁Less than 0.1* (2 ⁿ-1), S ₁/ S ₂Greater than 1.3, and S ₁₁Greater than S ₁₂, in operation 74, determine characteristic value P ₁Whether less than 0.1* (2 ⁿ-1), S ₁/ S ₂Whether greater than 1.2, and R ₁Whether greater than R ₂

If determine characteristic value P ₁Less than 0.1* (2 ⁿ-1), S ₁/ S ₂Greater than 1.2, and R ₁Greater than R ₂, in operation 75, coefficient a is detected as 0.5.

If do not determine characteristic value P ₁Less than 0.1* (2 ⁿ-1), S ₁/ S ₂Greater than 1.2, and R ₁Greater than R ₂, in operation 76, determine S ₁₁/ S ₁₂Whether greater than 1.3.

If determine S ₁₁/ S ₁₂Greater than 1.3, in operation 77, coefficient a is detected as 0.3.

If determine S ₁₁/ S ₁₂Be equal to or less than 1.3, in operation 78, coefficient a is detected as 0.

If in operation 71, determine characteristic value P ₂Be equal to or less than 0.98* (2 ⁿ-1), in operation 79, determines characteristic value P ₁Whether less than 0.1* (2 ⁿ-1), and S ₁/ S ₂Whether greater than 1.3 or S ₁₁Whether greater than S ₁₂

If determine characteristic value P ₁Less than 0.1* (2 ⁿ-1), and S ₁/ S ₂Greater than 1.3 or S ₁₁Greater than S ₁₂, in operation 80, coefficient a is detected as 0.5.

If do not determine characteristic value P ₁Less than 0.1* (2 ⁿ-1), and S ₁/ S ₂Greater than 1.3 or S ₁₁Greater than S ₁₂, in operation 81, determine characteristic value P ₁Whether less than 0.1* (2 ⁿ-1), and S ₁/ S ₂Whether greater than 1.2 or R ₁Greater than R ₂

If determine characteristic value P ₁Less than 0.1* (2 ⁿ-1), and S ₁/ S ₂Greater than 1.2 or R ₁Greater than R ₂, in operation 82, coefficient a is detected as 0.3.

If do not determine characteristic value P ₁Less than 0.1* (2 ⁿ-1), and S ₁/ S ₂Greater than 1.2 or R ₁Greater than R ₂, in operation 83, coefficient a is detected as 0.

If in operation 70, do not determine characteristic value R ₃Greater than characteristic value R ₂With characteristic value R ₁, in operation 84, determine characteristic value R ₁Whether greater than characteristic value R ₂With characteristic value R ₃

If determine characteristic value R ₁Greater than characteristic value R ₂With characteristic value R ₃, in operation 85, determine characteristic value S ₁₁Whether greater than characteristic value S ₁₂

If do not determine characteristic value S ₁₁Greater than characteristic value S ₁₂, in operation 86, coefficient a is detected as 0.

If determine characteristic value S ₁₁Greater than characteristic value S ₁₂, in operation 87, determine characteristic value P ₁Whether less than 0.1* (2 ⁿ-1), and S ₁/ S ₂Whether greater than 1.3.

If determine characteristic value P ₁Less than 0.1* (2 ⁿ-1), and S ₁/ S ₂Greater than 1.3, in operation 88, coefficient a is detected as 0.7.

If do not determine characteristic value P ₁Less than 0.1* (2 ⁿ-1) and S ₁/ S ₂Greater than 1.3, in operation 89, determine characteristic value S ₁Whether greater than characteristic value S ₂With characteristic value S ₃

If determine characteristic value S ₁Greater than characteristic value S ₂With characteristic value S ₃, in operation 90, coefficient a is detected as 0.5.

If do not determine characteristic value S ₁Greater than characteristic value S ₂With characteristic value S ₃, in operation 91, coefficient a is detected as 0.3.

If in operation 84, do not determine characteristic value R ₁Greater than characteristic value R ₂With characteristic value R ₃, in operation 92, determine S ₁₁/ S ₁₂, S ₁/ S ₂And S ₃/ S ₁In each whether greater than 1.3.

If determine S ₁₁/ S ₁₂, S ₁/ S ₂And S ₃/ S ₁In each greater than 1.3, the operation 93 in, coefficient a is detected as 0.3.

If do not determine S ₁₁/ S ₁₂, S ₁/ S ₂And S ₃/ S ₁In each greater than 1.3, the operation 94 in, coefficient a is detected as 0.

The operation sequence of above-described detection coefficient a and definite coefficient value only is illustrative.Utilize detected characteristic value can realize each one exemplary embodiment of the present invention, and the coefficient value of determining is not limited to above-mentioned numerical value and can be determined by different way.

Return with reference to the accompanying drawings 4, in operation 54, a RGB color coordinate system of image is converted into the color coordinate system of expression brightness.Just, the RGB color coordinate system of image is converted into HSV (tone, saturation, numerical value) color coordinate system, YcbCr color coordinate system, in Lab color coordinate system and the XYZ color coordinate system one.In each one exemplary embodiment of the present invention, the RGB color coordinate system of image can be converted into HSV color coordinate system.According to the image that is converted into HSV color coordinate system, can detect numerical value corresponding to a brightness value.

In operation 56, carry out border preservation (edge preserving) noise filtering to the image that is converted into the color coordinate system that represents brightness.

By to the capable filtering of image and subsequently based on the row filtering of row filtering result, perhaps to the row filtering of image and come border preservation noise filtering is carried out in the capable filtering of image based on row filtering result subsequently to image.

Utilize formula 9 and 10 to carry out the border and preserve noise filtering.

$V_f^{*}(r,c)=\frac{\underset{j=-s/2}{\overset{s/2}{\mathrm{\&Sigma;}}}V(r,c+j)*u\left(j\right)*w(V(r,c+j),V(r,c))}{\underset{j=-s/2}{\overset{s/2}{\mathrm{\&Sigma;}}}u\left(j\right)*w(V(r,c+j),V(r,c))}---\left(9\right)$

In formula 9, V (r, the c) brightness of expression original image, V _f ^*(r, c) expression is by (S represents a size that is used for the window of filtering, u (z) representation formula for r, a filtering in row and column c) and the result that obtains to image V $u\left(z\right)=\exp [-\frac{{\mathrm{<figure-callout\; id="2"\; label="z"\; filenames="A20061009986700251.png"\; state="\{\{state\}\}">z</figure-callout>}}^2}{2{\mathrm{\&sigma;}}_D^2}],$ W (x, y) representation formula $w(x,y)=\exp [-\frac{{(x-y)}^2}{2{\mathrm{\&sigma;}}_R^2}],$ And σ _DAnd σ _RThe expression predetermined constant.For example, can be set to σ _D=5 and σ _R=0.1* (2 ⁿ-1).

$V_f(r,c)=\frac{\underset{i=-s/2}{\overset{s/2}{\mathrm{\&Sigma;}}}V(r+i,c)*u\left(i\right)*w(V_f^{*}(r+i,c),V_f^{*}(r,c))}{\underset{i=-s/2}{\overset{s/2}{\mathrm{\&Sigma;}}}u\left(i\right)*w(V_f^{*}(r+i,c),V_f^{*}(r,c))}---\left(10\right)$

In formula 10, V _f(r, c) expression utilizes filtering V as a result _f ^*(r, c) result who obtains by one of them filtering at least to unfiltered row or column.For example, if utilize the V of the row of formula 9 detected image _f ^*(r c), utilizes the V of the row of formula 10 detected image _f(r, c).On the contrary, if utilize the V of the row of formula 9 detected image _f ^*(r c), then utilizes the V of the row of formula 10 detected image _f(r, c).

Accompanying drawing 7A-D illustrates the image that obtains by the operation of carrying out accompanying drawing 4.Accompanying drawing 7A illustrates original image, and accompanying drawing 7B illustrates by carrying out the result that noise filtering obtains is preserved on the border of original image.As shown in accompanying drawing 7B, owing to preserve noise filtering result this image blurring (blurred) with the border.

Owing to be by to the capable filtering of image and come row filtering based on capable filtering result subsequently to image,, can reduce amount of calculation with simultaneously the method for the row and column filtering of image being compared perhaps to the row filtering of image and come that based on row filtering result thereby the border is carried out in the capable filtering of image subsequently and preserve noise filtering.

In operation 58, go out the image that the contrast of specific region has wherein been regulated by having carried out the image detection of preserving noise filtering in the border.

In order to detect the image that the contrast of specific region has wherein been regulated, adopt formula 11.

D＝f(V _f，V)＝(2 ⁿ-1)*k*V/(V _f+b) (11)

In formula 11, D represents the image of contrast through regulating, and k and b represent the constant be scheduled to.For example, can be set at k=1.2 and b=3* σ _RBy using formula 11, the specific region of adjusting original image is the contrast of low brightness area for example.

Accompanying drawing 7C illustrates and utilizes formula 11 to preserve the image of the image of noise filtering by the contrast acquisition of adjusting specific region according to wherein having carried out the border.As shown in the accompanying drawing 7C,, improve the contrast of the dark area of filtering image by regulating for example contrast of dark area of specific region.

In operation 60, utilize contrast to regulate the brightness of specific region through the image of regulating.

In order to regulate the brightness of specific region, use formula 12.

V′＝f _W(V+a*(2 ⁿ-1-V _f)*f _b(D-V)) (12)

In formula 12, V ' represents the image that the wherein brightness of specific region has been regulated, and a represents to utilize characteristic value detected coefficient in operation 52, f _w(x) expression $f_w\left(x\right)=\{\frac{x:x\&le;(2^n-1)}{2^n-1:x>(2^n-1)},$ And f _b(x) expression $f_b\left(x\right)=\{\frac{x/(2^n-1):x\&GreaterEqual;0}{0:x<0}.$

Accompanying drawing 7D illustrates the image that the wherein brightness of specific region has been regulated, and it utilizes formula 12 to obtain from the image of contrast through regulating.As shown in the accompanying drawing 7D, be the brightness of dark area by regulating the specific region, can obtain the image that the brightness of the dark area of comparing with original image has been enhanced.

Accompanying drawing 8 is one and is used for according to the present invention and the schematic diagram of the brightness of the specific region of prior art movement images.1. mark is a curve chart according to the illuminometer aspect prime number quantitative changeization of original image, 2. mark is a curve chart according to the illuminometer aspect prime number quantitative changeization of image, this image is obtaining in a usual manner of all being raised of the brightness of wherein image, and 3. mark be a curve chart according to the illuminometer aspect prime number quantitative changeization of image, and this image obtains according to one exemplary embodiment of the present invention.As shown in Figure 8, in a usual manner, by all being heightened, the brightness of image can make the picture quality of high-brightness region of original image reduce, and according to one exemplary embodiment of the present invention, utilize said method, by only regulating the brightness of dark area, can prevent the reduction of the picture quality of relative high-brightness region, and can improve the picture quality of dark area simultaneously.

In operation 62, the color coordinate system of the image of brightness through regulating is converted into RGB color coordinate system.Just, be converted into color coordinate system that represents brightness and the image that its brightness has been regulated and be converted into RGB color coordinate system so that the people can discern actual image.

Can be written as code/instructions/programs and can carry out the general purpose digital computer of this code/instructions/programs and realize according to the method for brightness that is used for regulating image of one exemplary embodiment of the present invention by utilizing computer-readable to be recorded in medium.The example of these computer fetch mediums comprises magnetic recording medium (for example, ROM, floppy disk, hard disk etc.), optical recording media (for example, CD-ROM, perhaps DVD), and the medium of carrier wave (for example passing through Internet transmission) for example.Computer-readable recording medium can also distribute by the network that is connected with computer system and make the embodied on computer readable code to be stored and to carry out in the mode that distributes.In addition, be used to realize functional programs of the present invention, code and code snippet can easily be explained by the skilled programmer in field under the present invention.

Now, the device that is used to regulate image brightness according to embodiments of the invention is described with reference to the drawings.

Accompanying drawing 9 is the block diagrams of device that are used to regulate image brightness according to one exemplary embodiment of the present invention.With reference to the accompanying drawings 9, this device comprises 120, one brightness regulators 130 of 110, one contrast adjustment devices of 100, one verifiers of a distributed intelligence detector.

This distributed intelligence detector 100 detects by the distributed number information of the pixel brightness of the image of input terminal IN1 input and with testing result and outputs to verifier 110 and brightness regulator 130.Distributed number information is the information about the quantity of the pixel with same brightness, and it comes the pixel of classified image to obtain by the brightness according to each pixel.For example, it is very suitable showing distributed number information by histogram.In addition, when receiving the image of a contrast through regulating from contrast adjustment device 120, distributed intelligence detector 100 detects the distributed number information of the image of contrasts through regulating.

Verifier 110 need to determine whether the contrast adjustment of image and output to determine that the result gives contrast adjustment device 120.

Definite as a result the time when what receive the contrast adjustment that needs image from verifier 110, contrast adjustment device 120 is regulated the contrast of images and is regulated results to 100 outputs of distributed intelligence detector.

Accompanying drawing 10 is block diagrams of the contrast adjustment device 120 of accompanying drawing 9.With reference to the accompanying drawings 10, contrast adjustment device 120 comprises a reference brightness value detector 200 and a RGB adjuster 210.

Reference brightness value adjuster 200 from image, detect an indication low-light level the first reference brightness value and the second reference brightness value and the output testing result of an indication high brightness give RGB adjuster 210.

When corresponding at its pixel number greater than the brightness value of the minimum pixel number in the brightness value of first threshold quantity during less than first luminance threshold, reference brightness value detector 200 detects brightness value less than the first threshold brightness value as the first reference brightness value.In addition, when corresponding at its pixel number greater than the brightness value of the maximum pixel quantity in the brightness value of second number of thresholds during greater than the second threshold luminance value, reference brightness value detector 200 detects brightness value greater than the second threshold luminance value as the second reference brightness value.In order to detect the first reference brightness value and the second reference brightness value, reference brightness value detector 200 utilizes above-mentioned formula 1 and 2 to carry out these operations.

RGB adjuster 210 utilizes the detected first and second reference brightness values to regulate R, G and the B of image.Just, RGB adjuster 210 utilizes above-mentioned formula 3 to regulate the R of image, G and B.

Return with reference to the accompanying drawings 9, brightness regulator 130 utilizes the brightness value of regulating the image of a specific region according to the characteristic value of the distribution character of detected distributed number information.

Accompanying drawing 11 is block diagrams of the brightness regulator 130 of an accompanying drawing 9.With reference to the accompanying drawings 11, this brightness regulator 130 comprises a characteristic value detector 300, a coefficient detector 310, one first color coordinate system transducer 320, a filter unit 330,340, one specific region brightness regulators 350 of a specific region contrast adjustment device, and one second color coordinate system transducer 360.

Characteristic value detector 300 comes detected characteristics value and output detected characteristics value to give coefficient detector 310 according to the distribution character that detects distributed number information.

The distributed number information that characteristic value detector 300 will be expressed as histogram based on brightness is divided into a plurality of zones and detects the characteristic value of each cut zone.For example, it is one first cut zone that characteristic value detector 300 is cut apart a plurality of zones, one second cut zone and one the 3rd cut zone.Characteristic value detector 300 utilizes formula 4 to 8 detected characteristics values.

Parametric detector 310 utilizes detected characteristic value to detect to be used to the coefficient of the brightness of regulating the specific region and exports detected coefficient and give the first color coordinate system transducer 320.Coefficient detector 310 comes detection coefficient according to detected characteristic value by mutual more detected characteristic value, as shown in Figure 6.

The first color coordinate system transducer 320 is converted to a RGB color coordinate system of image the color coordinate system of an expression brightness and exports transformation results to filter unit 330.The first color coordinate system transducer 320 is converted to HSV color coordinate system, YCbCr color coordinate system, Lab color coordinate system with the RGB color coordinate system of image, and in the XYZ color coordinate system one.For example, the first color coordinate system transducer 320 is converted to HSV color coordinate system with the RGB color coordinate system of image.

The border of filter unit 330 carries out image is preserved noise filtering and is exported the filtering results to specific region contrast adjustment device 340.

The capable filtering of 330 pairs of images of filter unit and subsequently according to the row filtering of row filtering result to image is perhaps to the row filtering of image and subsequently according to the result of the row filtering capable filtering to image.Filter unit 330 utilizes the border of above-mentioned formula 9 and 10 carries out image to preserve noise filtering.As shown in the accompanying drawing 7B, come blurred picture by carrying out border preservation noise filtering.

Specific region contrast adjustment device 340 detects image that the contrast of specific region regulated and to specific region brightness regulator 350 output testing results from the image of having carried out the border and preserving noise filtering.The contrast that specific region contrast adjustment device 340 utilizes above-mentioned formula 11 to regulate the specific region.As shown in the accompanying drawing 7C, be the contrast of dark area by regulating the specific region, the contrast of the dark area of filtering image is improved.

Specific region brightness regulator 350 utilizes contrast to regulate the brightness of specific region and will regulate the result and export to the second color coordinate system transducer 360 through the image of regulating.Specific region brightness regulator 350 utilizes above-mentioned formula 12 to regulate the brightness of specific region.As shown in the accompanying drawing 7D, be the brightness of dark area by regulating the specific region, can obtain the image that the brightness of the dark area of comparing with original image has improved.

The second color coordinate system transducer 360 is converted to RGB color coordinate system with the color coordinate system of the image of brightness through regulating.Just, the second color coordinate system transducer 360 will be converted into the color coordinate system of expression brightness and the image transitions that its brightness has been regulated is a RGB color coordinate system, so that the people can discern real image.

As mentioned above, at the method and apparatus of brightness that is used for regulating image according to one exemplary embodiment of the present invention, by regulating the brightness of specific region (for example dark area), can stop the reduction of the picture quality of a relative high-brightness region, and improve the picture quality of dark area simultaneously.

Though the present invention has been carried out detailed expression and description with reference to one exemplary embodiment of the present invention, but those having ordinary skill in the art will appreciate that, under the situation that does not exceed the spirit and scope of the present invention that define as the claim that is proposed, can make various variations in form and details.

Claims (20)

1, a kind of method of regulating image brightness, this method comprises:

Detect the distributed number information of pixel brightness; With

Distribution character according to detected distributed number information utilizes characteristic value to regulate the brightness of the specific region of image.

2, the method for claim 1 further comprises:

Need to determine whether the contrast adjustment of image; With

The contrast adjustment of image if desired, the contrast of regulating image.

3, the method for claim 2, wherein the contrast adjustment of image comprises:

Detect first reference brightness value of an indicating image low-light level and the second reference brightness value of an indicating image high brightness;

Utilize the detected first and second reference brightness values to regulate the R of image, G and B; With

Detect the distributed number information of the image of RGB through regulating.

4, the method for claim 3, wherein, at the R that regulates image, among G and the B, formula is regulated R, G and the B of image below utilizing

R′(r，c)＝(2 ⁿ-1)*(R(r，c)-low)/(high-low)

G′(r，c)＝(2 ⁿ-1)*(G(r，c)-low)/(high-low)

B′(r，c)＝(2 ⁿ-1)*(R(r，c)-low)/(high-low)

Wherein, the coordinate of each pixel in r and the c presentation video, n are represented the constant of the quantity of an indication brightness degree, and low represents the first reference brightness value, and high represents the second reference brightness value.

5, the process of claim 1 wherein that the brightness regulation of specific region of image comprises:

The detected characteristics value;

Utilize detected characteristic value to detect a coefficient that is used to regulate the brightness of specific region;

The RGB color coordinate system of image is converted to the color coordinate system of an expression brightness;

The image of conversion is carried out the border preserve noise filtering;

The image that the contrast of its specific region of detection has been regulated from the image of having carried out border preservation noise filtering;

Utilize contrast to regulate the brightness of specific region through the image of regulating; With

The color coordinate system of the image of brightness through regulating is converted to RGB color coordinate system.

6, the method for claim 5, wherein in the detection of characteristic value, the distributed number information that will be expressed as histogram based on brightness is divided into a plurality of zones, and detects the characteristic value of each cut zone.

7, the method for claim 6, wherein a plurality of zones are split into one first cut zone, one second cut zone and one the 3rd cut zone.

8, the method for claim 7, wherein in the detection of characteristic value, formula comes the detected characteristics value below utilizing:

$\mathrm{gm}=\underset{i\&Element;[0,2^n-1]}{\max \left(H\left(i\right)\right)}$

Wherein, H (i) expression is corresponding to the pixel number of brightness value i, and gm is illustrated in corresponding to the maximum pixel quantity in the pixel number of the brightness value of image, and n represents to indicate the constant of the quantity of brightness degree;

$S_1=\underset{\mathrm{first}}{\mathrm{\&Sigma;}}H\left(i\right)/(M*N),S_2=\underset{\sec \mathrm{ond}}{\mathrm{\&Sigma;}}H\left(i\right)/(M*N),S_3=\underset{\mathrm{third}}{\mathrm{\&Sigma;}}H\left(i\right)/(M*N)$

Wherein, S ₁Expression is corresponding to the pixel number ratio of the brightness of first cut zone in the image, S ₂Expression is corresponding to the pixel number ratio of the brightness of second cut zone in the image, S ₃Expression is corresponding to the pixel number ratio of the brightness of the 3rd cut zone in the image, and the size of M*N presentation video (resolution);

$S_{11}=\underset{i=0}{\overset{\{(2^n-1)/6\}-1}{\mathrm{\&Sigma;}}}H\left(i\right)/(M*N),S_{12}=\underset{i=(2^n-1)/6}{\overset{(2^n-1)/3}{\mathrm{\&Sigma;}}}H\left(i\right)/(M*N),$

Wherein when the first area is split into first subregion and second subregion, S ₁₁Expression is corresponding to the pixel number ratio of the brightness of first subregion in the image, and S ₁₂Expression is corresponding to the pixel number ratio of the brightness of second subregion in the image;

$R_1=\underset{\mathrm{first}}{\max \left(H\left(i\right)\right)/\mathrm{gm}},R_2=\underset{\sec \mathrm{ond}}{\max \left(H\left(i\right)\right)/\mathrm{gm}},R_3=\underset{\mathrm{third}}{\max }\left(H\left(i\right)\right)/\mathrm{gm},$

Wherein, R ₁Expression is corresponding to maximum pixel quantity in the pixel number of the brightness value in first cut zone and ratio corresponding to the maximum pixel quantity in the pixel number of the brightness value of image, R ₂Expression is corresponding to maximum pixel quantity in the pixel number of the brightness value in second cut zone and ratio corresponding to the maximum pixel quantity in the pixel number of the brightness value of image, and R ₃Expression is corresponding to maximum pixel quantity in the pixel number of the brightness value in the 3rd cut zone and ratio corresponding to the maximum pixel quantity in the pixel number of the brightness value of image; With

P ₁＝h/(2 ⁿ-1)，P ₂＝l/(2 ⁿ-1)

Wherein, h represents to have the brightness value of the maximum pixel quantity in first cut zone, P ₁Expression h and 2 ⁿThe ratio of grade, brightness value of 1 expression, P with the maximum pixel quantity in the 3rd cut zone ₂Expression 1 and 2 ⁿThe ratio of grade.

9, the method for claim 5, wherein be converted in the color coordinate system of expression brightness in RGB color coordinate system with image, the RGB color coordinate system of image is converted into HSV (tone, saturation, numerical value) color coordinate system, YCbCr color coordinate system, Lab color coordinate system, and in the XYZ color coordinate system one.

10, the method for claim 5 is wherein preserved in the noise filtering carrying out the border, to the capable filtering of image and subsequently according to the row filtering of capable filtering result to image, perhaps to the row filtering of image and subsequently according to the result of the row filtering capable filtering to image.

11, the method for claim 10, wherein in carrying out border preservation noise filtering, formula is carried out the border and is preserved noise filtering below utilizing;

$V_f^{*}(r,c)=\frac{\underset{j=-s/2}{\overset{s/2}{\mathrm{\&Sigma;}}}V(r,c+j)*u\left(j\right)*w(V(r,c+j),V(r,c))}{\underset{j=-s/2}{\overset{s/2}{\mathrm{\&Sigma;}}}u\left(j\right)*w(V(r,c+j),V(r,c))}$

Wherein, V (r, the c) brightness of expression original image, V _f ^*(r, c) expression is by (S represents a window size that is used for filtering, u (z) representation formula for r, a filtering in row or column c) and obtain the result to image V $u\left(z\right)=\exp [-\frac{z^2}{2{\mathrm{\&sigma;}}_D^2}],$ W (x, y) representation formula $w(x,y)=\exp [-\frac{{(x-y)}^2}{2{\mathrm{\&sigma;}}_R^2}],$ And σ _DAnd σ _RThe expression predetermined constant; And

$V_f(r,c)=\frac{\underset{j=-s/2}{\overset{s/2}{\mathrm{\&Sigma;}}}{V}_f^{*}(r+i,c)*u\left(j\right)*w(V_f^{*}(r+i,c),V_f^{*}(r,c))}{\underset{j=-s/2}{\overset{s/2}{\mathrm{\&Sigma;}}}u\left(j\right)*w(V_f^{*}(r+i,c),V_f^{*}(r,c))}$

Wherein, V _f(r, c) expression utilizes filtering V as a result _f ^*(r, c) result by unfiltered row or column filtering is obtained.

12, the method for claim 5, wherein in the image that the contrast that detects its specific region has been regulated, formula detects the image that the contrast of its specific region has been regulated below adopting,

D＝f(V _f，V)＝(2 ⁿ-1)*k*V/(V _f+b)

In formula, D represents the image of a contrast through regulating, and k and b represent predetermined constant.

13, the method for claim 5, wherein in the brightness of regulating the specific region, formula is regulated the brightness of specific region below utilizing

V′＝f _w(V+a*(2 ⁿ-1-V _f)*f _b(D-V))

In formula, the image that the brightness of its specific region of V ' expression has been regulated, a represents to utilize the detected coefficient of characteristic value, f _w(x) expression $f_w\left(x\right)=\{\frac{x:x\&le;(2^n-1)}{2^n-1:x>(2^n-1)},$ And f _b(x) expression

$f_b\left(x\right)=\{\frac{x/(2^n-1):x\&GreaterEqual;0}{0:x<0}.$

14, a kind of computer-readable recording medium of having stored the computer-readable program of the method that is used for one of enforcement of rights requirement 1 to 5.

15, a kind of device that is used to regulate image brightness, this device comprises:

A distributed intelligence detector that detects the distributed number information of pixel brightness; With

Distribution character according to detected distributed number information utilizes characteristic value to regulate the brightness regulator of brightness of the specific region of image.

16, the device of claim 15 further comprises:

A verifier that need to determine whether the contrast adjustment of image; With

Regulate the contrast adjustment device of the contrast of image if desired during the contrast adjustment of image for one.

17, the device of claim 16, contrast adjustment device wherein comprises:

A reference brightness value detector is used to detect first reference brightness value of an indicating image low-light level and the second reference brightness value of an indicating image high brightness; With

A RGB adjuster is used to utilize the detected first and second reference brightness values to regulate the R of image, G and B,

Wherein the distributed intelligence detector detects the distributed number information of the image of RGB through regulating.

18, the device of claim 15, wherein brightness regulator comprises:

A characteristic value detector is used for the detected characteristics value;

A coefficient detector utilizes detected characteristic value to detect to be used to the coefficient of the brightness of regulating the specific region;

One first color coordinate system transducer is used for a RGB color coordinate system of image is converted to a color coordinate system that represents brightness;

A filter unit, noise filtering is preserved on the border that is used to carry out the image of conversion;

A specific region contrast adjustment device, the image that the contrast of its specific region of detection has been regulated from the image of having carried out border preservation noise filtering;

A specific region brightness regulator utilizes contrast to regulate the brightness of specific region through the image of regulating; With

One second color coordinate system transducer is used for the color coordinate system of the image of brightness through regulating is converted to RGB color coordinate system.

19, the device of claim 18, wherein the characteristic value detector is divided into a plurality of zones based on the distributed number information that brightness will be expressed as histogram, and detects the characteristic value of each cut zone.

20, the device of claim 18, filter unit wherein to the capable filtering of image and subsequently according to the result of row filtering to the row filtering of image, perhaps to the row filtering of image and subsequently according to the result of row filtering capable filtering to image.

Images