CN101453558A - Video image contrast improving method - Google Patents

Abstract

The invention provides a method for improving the contrast of a video image. The method comprises: performing histogram equalization processing on the image first, being capable of identifying whether pixels of the current image are concentrated in low gray scale regions, and judging whether further processing is required after histogram equalization, namely when the ratio of the number of the pixels of the current image in the low gray scale regions to the total pixel number of the image is more than a set proportional threshold R, the pixels of the current image are concentrated in the low gray scale regions and further processing is required after histogram equalization of the image; and when the ratio of the number of the pixels of the current image in the low gray scale regions to the total pixel number of the image is less than or equal to the set proportional threshold R, the pixels of the current image are not concentrated in the low gray scale regions and an image gray value after histogram equalization is only required to be outputted directly. The method reduces the computation complexity and computation when simultaneously improving the contrast of the video image and enriching the stereovision of the image.

Description

A kind of video image contrast improvement method

Technical field

The invention belongs to field of video image processing, the method that particularly a kind of video image contrast is improved.

Background technology

LCD TV is with respect to traditional C RT and plasm TV, congenital weakness with contrast deficiency, though the contrast of a lot of LCD TV producers propaganda reaches incredible numeral, but its index that adopts is instantaneous maximum-contrast, these data can not really embody the contrast of LCD TV image, actual visual effect is still undesirable, therefore improves the emphasis that the video image contrast is industry research.

Histogram equalization is the present common method of improving the image/video contrast, and its principle is the details expressive ability that strengthens image by the dynamic range of preliminary treatment increase image.This method belongs to the method that the overall situation is handled, and all pictures are all carried out expansion of gradation, but gray scale is being presented the effect that two ends distribute, overexposure can appear in the low gray areas of while image when having the image of more pixel distribution to handle.In order to overcome this defective, generally need further process as the histogram matching treatment, but the computing of histogram coupling is equivalent to the computing of a histogram equalization, and operand is doubled at least image, complexity increases greatly.

Summary of the invention

The object of the present invention is to provide a kind of method of improving the video image contrast, with rich image stereovision and reduction computational complexity.

The object of the present invention is achieved like this: a kind of video image contrast improvement method, described method is at first carried out histogram equalization to image and is handled, whether described method can intensive in low gray areas according to identification present image pixel, and whether judge needs further processing behind histogram equalization; When promptly the ratio that accounts for this image total pixel number at the pixel count of low gray areas when present image is greater than the proportion threshold value R that sets, show that then the present image pixel-intensive in low gray areas, needs further processing after image is carried out histogram equalization; The ratio that accounts for the total pixel of this image in the pixel of low gray areas when present image is during smaller or equal to the proportion threshold value R that sets, and it is intensive in low gray areas to show that then the present image pixel does not have, and only needs directly to export the gradation of image value behind the histogram equalization.

Described present image obtains in the following manner at the pixel count of low gray areas: set gray level thresholding c, rule of thumb set the low gray value M of image _Min, calculate at low gray value empirical value M _MinC gray scale accumulated value of carrying out statistics with histogram is gained up and down.

The span of described proportion threshold value R is [0.3,0.5], and the span of described c is [10,30].

Further, described further processing comprises the histogram matching treatment.

Method of the present invention is according to the statistics of grey level histogram, and whether recognition image is intensive at low gray area, promptly image is carried out light and shade and judges, when image is not in low gray area, only needs carry out histogram equalization to image; Otherwise, behind histogram equalization, need to carry out again the histogram coupling, thereby when improving video image contrast, rich image stereovision, reduced computational complexity, reduced operand.

Embodiment

Below will be described in further detail a kind of video image contrast improvement method of the present invention.

A kind of video image contrast improvement method of the present invention, the employing histogram equalizing method is realized, pixel for fear of low gray areas the overexposure phenomenon occurs through behind the histogram equalizing method, and the image after this method is handled histogram equalization carries out the histogram matching treatment; But histogram matching treatment operand is big, the computing complexity.Under the prerequisite that guarantees the improvement of video image contrast, reduce computational complexity as much as possible, whether intensive at low gray area, only the image that is in low gray areas is done the histogram matching treatment then if need in processing procedure, identify the present image pixel.When promptly the ratio that accounts for this image total pixel number at the pixel count of low gray areas when present image is greater than the proportion threshold value R that sets, show that then the present image pixel-intensive in low gray areas, needs further processing after image is carried out histogram equalization; The ratio that accounts for the total pixel of this image in the pixel of low gray areas when present image is during smaller or equal to the proportion threshold value R that sets, and it is intensive in low gray areas to show that then the present image pixel does not have, and only needs directly to export the gradation of image value behind the histogram equalization.

For characteristics of the present invention and technique effect are described better, described in detail especially exemplified by following specific embodiment.

According to a kind of video image contrast improvement method provided by the invention, yuv format image file degree of comparing is improved processing, its concrete steps are as follows:

Step 1, at first read in entire image, calculate the total pixel number of this image, be designated as S _Um, import the empirical value M that this image hangs down gray value _Min=30;

Step 2, then statistical picture grey level histogram H[256]; And histogram made normalized, obtain probability density function:

P(r _i)＝H[i]/S _um i＝0，1，2...L-1

Wherein L is the number of this gradation of image grade.

Step 3, calculating probability density function P (r _i) corresponding cumulative distribution function s _k, utilize T (r) to carry out conversion, select following transformation law:

$\begin{array}{cc}{s}_k=T\left(r_k\right)=\underset{i=0}{\overset{k}{\mathrm{\Σ}}}P\left(r_i\right)& i=\mathrm{0,1,2}...L-1\end{array}$

Step 4, calculating are 30 o'clock at gray level thresholding c, and present image is at the pixel count of low gray areas

Rule of thumb the span of R is [0.3,0.5], and setting the R value in the present embodiment is 0.4, then present image is carried out light and shade and judges, promptly judges With the size of R value, wherein,, also need image is carried out the histogram matching treatment if the former greater than the latter, illustrates that then present image is in inclined to one side dark space, promptly carry out step 5; If the former smaller or equal to the latter, illustrates that then present image is not to be in inclined to one side dark space, do not need to carry out the histogram matching treatment, promptly carry out step 6;

Step 5, setting P _z(z _i) being output image expectation histogram, z is the gray scale of output image.Calculate this probability density function p _z(z _i) cumulative distribution function G (z _k):

$\begin{array}{cc}G\left(z_k\right)=\underset{i=0}{\overset{k}{\mathrm{\Σ}}}{p}_z\left(z_i\right)=s_k& k=\mathrm{0,1,2}...L-1\end{array}$

Then the final output gray level value of pixel is:

z _k＝G ^-1[s _k]＝G ^-1[T(r _k)] k＝0，1，2...L-1

Step 6, the final output gray level value of calculating pixel are:

z _k＝INT[(g _max-g _min)T(r _k)+g _min+0.5] k＝0，1，2...P-1

G in the formula _Max, g _MinBe respectively gray scale maximum, minimum value in the original image, P is that histogram equalization is handled back tonal gradation sum, and k is the greyscale level of image.

Claims (6)

1. video image contrast improvement method, described method is at first carried out histogram equalization to image and is handled, whether it is characterized in that whether described method can intensive in low gray areas according to identification present image pixel, judging needs further processing behind histogram equalization; When promptly the ratio that accounts for this image total pixel number at the pixel count of low gray areas when present image is greater than the proportion threshold value R that sets, show that then the present image pixel-intensive in low gray areas, needs further processing after image is carried out histogram equalization; The ratio that accounts for the total pixel of this image in the pixel of low gray areas when present image is during smaller or equal to the proportion threshold value R that sets, and it is intensive in low gray areas to show that then the present image pixel does not have, and only needs directly to export the gradation of image value behind the histogram equalization.

2. video image contrast improvement method as claimed in claim 1, it is characterized in that, described present image obtains in the following manner at the pixel count of low gray areas: set gray level thresholding c, rule of thumb set the low gray value Mmin of image, calculate low gray value empirical value Mmin up and down c gray scale accumulated value of carrying out statistics with histogram be gained.

3. video image contrast improvement method as claimed in claim 2 is characterized in that the span of described gray level thresholding c is [10,30].

4. video image contrast improvement method as claimed in claim 1 is characterized in that the span of described proportion threshold value R is [0.3,0.5].

5. video image contrast improvement method as claimed in claim 4 is characterized in that the value of described proportion threshold value R is 0.4.

6. video image contrast improvement method as claimed in claim 1 is characterized in that described further processing comprises the histogram matching treatment.