CN101957988A - Method and device for obtaining probability distribution of image grey spots and white balance method and device - Google Patents

Abstract

Embodiments of the present invention provide a method and device for obtaining the probability distribution of image gray points, and a white balance method and device. Wherein, a method for obtaining the probability distribution of image grayscale points includes: taking a reference image with color patches and grayscale patches; establishing color points according to the color values of each pixel corresponding to the color patches in the reference image histogram, and according to the color value of each pixel point corresponding to the gray color block in the reference image, a gray point histogram is established; the color point histogram and the gray point histogram are equally divided into multiple intervals, according to The number of pixels in the corresponding intervals in the color point histogram and the gray point histogram is used to obtain the gray point probability distribution map of each pixel in the reference image. By obtaining the gray point probability distribution map of the reference image, the judgment of the gray point of the captured image is more accurate, and the accuracy of the white balance gain of the captured image is improved.

Description

Method and device for obtaining probability distribution of image gray points and white balance method and device

technical field

The present invention relates to the technical field of image processing, in particular to a method and device for obtaining the probability distribution of image gray points, and a white balance method and device.

Background technique

In the process of photography, the light emitted from the light source is reflected by the object, and the image sensor of the photographic equipment obtains the image of the object by measuring the reflected light.

It can be seen that the imaging of objects in photography is related to the lighting conditions and the objects themselves. For example, when a yellow light source shines light on a white wall, the reflected light also appears yellow to the image sensor because the white object reflects all the incident light. In this way, the imaging of the photographed object is inconsistent with what our human eyes observe, that is, the white balance is inaccurate. It is necessary to adjust the white balance to make the image obtained by photography consistent with the image observed by human eyes.

At present, the commonly used white balance method, for example, determines the gray point in the image by setting a fixed threshold. It should be known that the gray point refers to the pixel point with the same value of each color channel when the white balance is correct, such as the three primary color channels of red, green and blue (R, G, B) R=G=B. In this way, by adjusting the color channel value of the determined gray point to R=G=B, the gain value of each color channel of the image is obtained, and then the image is adjusted to achieve white balance.

In the course of realizing the present invention, the inventor finds that there are at least the following problems in the prior art:

Due to the inaccuracy of the white balance, the gray point deviates from a certain color, but the degree of deviation is determined by the lighting conditions and will not be fixed within a certain range. Therefore, a fixed threshold is used to determine the gray point. Poor accuracy and poor white balance.

Contents of the invention

Embodiments of the present invention provide a method and device for obtaining the probability distribution of image gray points, and a white balance method and device, which can accurately determine gray points and achieve good white balance effects.

A method for obtaining the probability distribution of image gray points, comprising:

Take a reference image with color patches and grayscale patches;

Establish a color point histogram according to the color values of each pixel corresponding to the color block in the reference image, and establish a gray scale according to the color values of each pixel corresponding to the gray color block in the reference image point histogram;

Dividing the color point histogram and the gray point histogram into a plurality of intervals respectively, according to the number of pixels in the corresponding intervals in the color point histogram and the gray point histogram, A gray point probability distribution diagram of each pixel point in the reference image is obtained.

A device for obtaining the probability distribution of image gray points, comprising:

An image acquisition unit, used for taking reference images with color patches and grayscale patches;

A histogram building unit, configured to create a color point histogram according to the color values of the pixels corresponding to the color blocks in the reference image, and to create a histogram of color points according to the values of the pixels corresponding to the gray color blocks in the reference image Color value, build gray point histogram;

a histogram analysis unit, configured to equally divide the color point histogram and the gray point histogram into multiple intervals, and obtain corresponding intervals in the color point histogram and the gray point histogram The number of pixels in

The gray point probability calculation unit is configured to obtain the gray point probability distribution map of the reference image according to the number of pixels in the color point histogram and the corresponding interval in the gray point histogram.

A white balance method comprising:

According to each color value of each pixel point in the captured image, and the gray-scale point probability distribution map of the reference image, the gray-scale point probability value of each pixel point in the captured image is obtained;

Obtaining an illumination estimate of the captured image according to the gray point probability value of each pixel in the captured image;

According to the illumination estimation of the captured image, the gain value of each color channel of the captured image is obtained to adjust the captured image to achieve white balance.

A white balance device, comprising:

A gray-scale point probability value acquisition unit, configured to obtain the gray-scale point probability value of each pixel point in the captured image according to each color value of each pixel point in the captured image and the gray-scale point probability distribution map of the reference image;

An illumination estimation unit, configured to obtain an illumination estimation of the captured image according to the gray point probability value of each pixel in the captured image;

The gain value obtaining unit is configured to obtain the gain value of each color channel of the captured image according to the illumination estimation of the captured image, so as to adjust the image to achieve white balance.

It can be seen from the technical solutions provided by the above-mentioned embodiments of the present invention that the present invention provides a statistical method for the probability distribution of gray points, which is used to count the probability distribution of gray points in the captured image, and based on the probability distribution of gray points According to the gray point probability distribution map of the reference image, the gray point probability value of each pixel point in the captured image is confirmed, and the white balance method is performed. Through the probability distribution of gray points, the judgment of gray points is more accurate, and the accuracy of white balance gain calculation is improved.

Description of drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the accompanying drawings that need to be used in the description of the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For Those of ordinary skill in the art can also obtain other drawings based on these drawings without any creative effort.

Fig. 1 is a schematic flow chart of a method for obtaining the probability distribution of image gray points according to an embodiment of the present invention;

2 is a schematic diagram of a color point histogram in a method for obtaining the probability distribution of image gray points according to an embodiment of the present invention;

3 is a schematic diagram of a gray point histogram in a method for obtaining the probability distribution of image gray points according to an embodiment of the present invention;

Fig. 4 is a schematic diagram of interval division in the color point histogram and the gray point histogram in the method for obtaining the probability distribution of image gray points in the embodiment of the present invention;

5 is a schematic diagram of a probability distribution diagram of gray points in a method for obtaining the probability distribution of image gray points according to an embodiment of the present invention;

Fig. 6 is a schematic diagram of a device for obtaining the probability distribution of image gray points according to an embodiment of the present invention;

FIG. 7 is a first schematic flow diagram of a white balance method according to an embodiment of the present invention;

FIG. 8 is a second schematic flow diagram of a white balance method according to an embodiment of the present invention;

FIG. 9 is a first schematic diagram of a white balance device according to an embodiment of the present invention;

Fig. 10 is a second schematic diagram of the composition of the white balance device according to the embodiment of the present invention;

FIG. 11 is a schematic flow chart of an application scenario of a white balance method according to an embodiment of the present invention.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

Based on the white balance method in the prior art, the accuracy of determining the gray point with a fixed threshold is poor, and the white balance effect is poor. Embodiments of the present invention provide a method and device for obtaining the probability distribution of image gray points, and a white balance method and device.

Under the known illumination conditions, the reference image is captured by shooting, the probability distribution of the gray points of the reference image is confirmed, and the probability distribution map of the gray points is obtained. Then, for the current image captured, the gray point probability value of each pixel can be obtained from the gray point probability distribution map to estimate the lighting situation of the current captured image, and the lighting estimation situation can be reflected by the value of the color channel . Then select the appropriate gain value of the color channel according to the lighting estimation situation, and adjust the current captured image to achieve the correct white balance.

It can be understood that a grayscale point refers to a pixel point with equal value of each color channel when the white balance is correct.

Lighting can include outdoor lighting or indoor lighting. Outdoor lighting includes: sunny sunlight, cloudy sunlight or shadows (for example, images captured under the shade of trees are shadow situations), etc. Indoor lighting includes: tungsten lighting, daylight lighting or fluorescent lighting Light, wait.

In the white balance method and device provided by the embodiments of the present invention, the process of adjusting the white balance can be completed through an automatic process, which belongs to the automatic white balance.

Embodiment one

As shown in Figure 1, an embodiment of the present invention provides a method for obtaining the probability distribution of image gray points, including:

Step 11, taking a reference image with color patches and grayscale patches.

Step 12, according to each color value of each pixel point corresponding to the color color block in the reference image, establish a color point histogram, and according to the color value of each pixel point corresponding to the gray color block in the reference image, establish a gray point histogram .

Step 13, the color point histogram and the grayscale point histogram are equally divided into a plurality of intervals respectively, and according to the number of pixels in the corresponding intervals in the color point histogram and the grayscale point histogram, obtain each Probability distribution map of grayscale points of pixels.

It can be known that each color value of each pixel in the image can be R, G, B in the RGB chromaticity space, or Cb, Cr in the Ycbcr chromaticity space, and so on. Usually, the color value of the pixel point of the image is known, and the color channel value of the image is also known.

Specifically, in step 11, optionally, the reference image is obtained by photographing a standard color palette with color patches and grayscale color patches. The standard color palette can be a standard 24 color palette, or other color Blocks in turn contain standard swatches of colored blocks.

As shown in Figures 2 and 3, in step 12, optionally, the two dimensions of the color point histogram and the gray point histogram may include:

R÷G and B÷G, or (R-G)÷G and (B-G)÷G of each pixel in the RGB chromaticity space.

Or, each pixel is in Cb and Cr in the YCbCr chromaticity space.

As shown in Figure 4, step 13 is to divide the color point histogram and the gray point histogram into multiple intervals according to the same division standard, and each interval corresponds to the number of pixels falling into the interval , the number of pixels is the value of the interval.

If the number of pixels in the interval in the gray point histogram is a, that is, the value of this interval is a, and the number of pixels in the corresponding interval in the color point histogram is b, that is, the value of this interval is b, then The probability value that the pixels in this interval are grayscale points is p=a÷(a+b), which refers to the probability that the color combination (such as RGB) corresponding to this interval may be grayscale points. In this way, a gray point probability distribution map of each pixel in the reference image (as shown in FIG. 4 ) can be obtained.

When (a+b)=0, then p=0. In addition, when the value of the pixel point is outside the expression range of the histogram, p=0 is also taken.

Optionally, step 11 of photographing a reference image with color patches and grayscale patches may be, under preset illumination, photographing a standard color palette with color patches and gray patches to obtain a reference image.

Preset lighting can include outdoor lighting or indoor lighting. Outdoor lighting includes: sunny daylight, cloudy daylight or shadow (for example, the image captured under the shade of a tree is the shadow situation), etc., and indoor lighting includes: tungsten lighting, daylight lighting or fluorescent lighting, etc.

In this way, the probability distribution map of the reference image obtained by outdoor lighting or indoor lighting can be stored respectively. When performing a shooting operation, the user can select the current shooting scene mode as preset lighting, such as indoor lighting or outdoor lighting. According to the gray point probability distribution map of the preset illumination of the selected scene mode, the automatic white balance of the image can be realized. For details, please refer to the following description.

In the same way, the probability distribution maps of reference images such as sunny daylight, cloudy daylight or shadow, tungsten light, daylight light or fluorescent light can be stored separately. When performing photography operations, the user can select the current shooting scene mode as preset Set the lighting, according to the gray point probability distribution map of the preset lighting of the selected scene mode, the automatic white balance of the image can be realized.

Embodiments of the present invention provide a method for obtaining the probability distribution of image gray points, and may further filter the probability distribution map obtained in step 14 to make the probability distribution map smoother.

It can be seen from the technical solutions provided by the above embodiments of the present invention that by obtaining the probability distribution map of the gray points of the reference image, the judgment of the gray points is more accurate. By referring to the gray point probability distribution of the image, the gray point probability of each point in the captured image is confirmed, and the illumination estimation and white balance are performed based on this, so there will be no gray point deviation and a better white balance effect.

Embodiment two

As shown in Figure 6, corresponding to the method for obtaining the probability distribution of image gray points in Embodiment 1, an embodiment of the present invention provides a device for obtaining the probability distribution of image gray points, including:

The image acquiring unit 61 is configured to capture a reference image having color patches and grayscale patches.

The histogram building unit 62 is used to establish a color point histogram according to the color values of each pixel point corresponding to the color color block in the reference image, and to establish a color point histogram according to the color values of each pixel point corresponding to the gray color block in the reference image. Histogram of grayscale points.

The histogram analysis unit 63 is configured to equally divide the color point histogram and the gray point histogram into a plurality of intervals, and obtain the number of pixels in corresponding intervals in the color point histogram and the gray point histogram.

The gray point probability calculation unit 64 is configured to obtain the gray point probability distribution diagram of the reference image according to the number of pixels in the corresponding intervals of the color point histogram and the gray point histogram.

Optionally, the image acquisition unit 61 can take a standard color palette with color patches and grayscale patches, the standard palette can be a standard 24 color palette, or other standard colors that contain both grayscale patches and color patches plate.

Optionally, the two dimensions of the color point histogram and the gray point histogram established by the histogram building unit 62 may include:

R÷G and B÷G, or (R-G)÷G and (B-G)÷G of each pixel in the RGB chromaticity space.

Or, each pixel is in Cb and Cr in the YCbCr chromaticity space.

The grayscale point probability calculation unit 64 is used for the value of the interval in the grayscale point histogram to be a, and the value of the corresponding interval in the color point histogram to be b, and the probability value that the pixel point in the interval is calculated as a grayscale point is p=a÷(a+b). When (a+b)=0, then p=0. When the value of the pixel is outside the expression range of the histogram, p=0 is also taken.

Optionally, the image acquisition unit 61 is also configured to capture a standard color palette with color patches and grayscale patches under preset illumination to obtain a reference image. Preset lighting includes: outdoor lighting or indoor lighting, outdoor lighting includes: sunny daylight, cloudy daylight or shadow, and indoor lighting includes: tungsten lighting, daylight lighting or fluorescent lighting.

An embodiment of the invention provides a device for obtaining a probability distribution map of gray-scale points of an image, and may further include: a filter (not shown in FIG. 6 ) for filtering the probability distribution map obtained by the gray-scale point probability calculation unit 64, Make the probability distribution plot smoother.

The image acquisition unit 61 may be a CCD (Charge Couple Device, charge-coupled device) or a CMOS (Complementary Metal-Oxide Semiconductor, complementary metal oxide semiconductor) photosensitive device. The image acquisition unit 61 outputs digital image information to the histogram analysis unit 63 .

It can be seen from the technical solutions provided by the above embodiments of the present invention that by obtaining the probability distribution map of the gray points of the reference image, the judgment of the gray points is more accurate. By referring to the gray point probability distribution of the image, the gray point probability of each point in the captured image is confirmed, and the illumination estimation and white balance are performed based on this, so there will be no gray point deviation and a better white balance effect.

Embodiment three

As shown in FIG. 7 , based on the method for obtaining the probability distribution of image gray points in the above embodiment, the embodiment of the present invention provides a white balance method, including:

Step 71. Obtain the grayscale point probability value of each pixel point in the captured image according to the color values of each pixel point in the captured image and the grayscale point probability distribution map of the reference image.

Step 72. According to the gray point probability value of each pixel in the captured image, obtain the illumination estimate of the captured image.

Step 73 , according to the illumination estimation of the captured image, obtain the gain value of each color channel of the captured image to adjust the captured image to achieve white balance.

In step 71, the photographing scene mode for obtaining the captured image is consistent with the preset photographing scene mode for obtaining the gray point probability distribution map of the reference image. If the lighting is all outdoor lighting or indoor lighting, details will not be described here.

As shown in Figure 8, optionally, according to each color value of each pixel in the captured image, step 71 of obtaining the gray-scale point probability value of each pixel in the captured image from the gray-scale point probability distribution map of the reference image, Can include:

Step 711, according to each color value of each pixel in the captured image, determine the corresponding position of each pixel in the gray point probability distribution map of the reference image;

Step 712 , according to the corresponding position of each pixel in the gray point probability distribution map of the reference image, obtain the gray point probability value corresponding to each pixel.

Specifically, if in the RGB chromaticity space,

Step 711 can be: according to the R, G, and B color values of each pixel in the RGB chromaticity space in the captured image, obtain R÷G and B÷G of each pixel, or obtain (R-G)÷G of each pixel G and (B-G)÷G, that is, the position of the horizontal and vertical axes of each pixel point in the gray point probability distribution map of the reference image, and then, it can be determined that each pixel point is in the gray point probability distribution map of the reference image specific location.

Step 712 may be: according to the corresponding position of each pixel point in the gray-scale point probability distribution map of the reference image, the gray-scale point probability value corresponding to the position may be obtained.

Or, if in YCbCr chromaticity space,

Step 711 can be: according to the Cb and Cr color values of each pixel in the YCbCr chromaticity space in the captured image, the position of the horizontal and vertical axes of each pixel in the gray point probability distribution diagram of the reference image is obtained, and then, it can be determined Find out the specific position of each pixel in the gray point probability distribution map of the reference image.

Step 712 may be: according to the corresponding position of each pixel point in the gray-scale point probability distribution map of the reference image, the gray-scale point probability value corresponding to the position may be obtained.

Optionally, the step 72 of obtaining the illumination estimation of the captured image according to the gray point probability value of each pixel in the captured image may include:

The gray point probability value of each pixel in the captured image is weighted to calculate the weighted average of each color value of all pixels, and the average value of each color channel of the captured image is obtained as the illumination estimation of the captured image.

Pgray(C_i)为像素点的灰度点概率值，C_i，k为像素点各颜色值，k为R或G或B，i为像素点个数编号。Specifically, in step 72, the illumination estimation calculation formula of the captured image is

Pgray(C _i ) is the probability value of the gray point of the pixel, C _{i, k} is the color value of each pixel, k is R or G or B, and i is the number of the pixel.

like, $L_R=\frac{\underset{i}{\mathrm{\Σ}}{P}_{\mathrm{gray}}\left(C_i\right)*C_{i,R}}{\underset{i}{\mathrm{\Σ}}1};$

${\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; G</span>}}<span\; class="notranslate">\; =</span>\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; gray</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; G</span>}}}{\underset{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ;</span>$

${\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; B</span>}}<span\; class="notranslate">\; =</span>\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; gray</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; B</span>}}}{\underset{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; .</span>$

得到拍摄图像R、G、B通道平均值L_R、L_G、L_B，即得到拍摄图像的光照估计结果。In the RGB chromaticity space, the formula

Obtain the average values L _R , L G , and L _B of the R, _G , and B channels of the captured image, that is, obtain the illumination estimation result of the captured image.

得到拍摄图像R、G、B通道平均值L_R、L_G、L_B，即得到拍摄图像的光照估计结果。Alternatively, in the YCbCr chromaticity space, YCbCr can be converted to RGB by the formula

Obtain the average values L _R , L G , and L _B of the R, _G , and B channels of the captured image, that is, obtain the illumination estimation result of the captured image.

Optionally, according to the illumination estimation of the captured image, the step 73 of obtaining the gain value of each color channel of the captured image to adjust the captured image to achieve white balance may include:

Step 731: Take the inverse of the average value of each color channel of the captured image to obtain the gain value of each color channel.

Step 732, realize the white balance adjustment of the captured image through the gain value of each color channel.

Specifically, in step 73, the R channel average value _LR is taken as an inverse number to obtain the gain value of the R channel for achieving white balance, that is, R _gain =1/L _R .

Taking the reciprocal of the average value L _G of the G channel, the gain value of the G channel for achieving white balance is obtained, that is, G _gain =1/L _G .

Taking the reciprocal of the average value L _B of the B channel, the gain value of the B channel for realizing white balance is obtained, that is, B _gain =1/L _B .

In this way, the white balance adjustment of the image can be realized by multiplying the value of each color channel by the gain value of each color channel.

It can be seen from the white balance method provided by the above-mentioned embodiments of the present invention that the gray point probability value of each pixel in the image can be obtained according to the gray point probability distribution map of the reference image, so that the judgment of the gray point is more accurate , thereby improving the accuracy of white balance gain calculation.

Moreover, the white balance method provided by the embodiments of the present invention, on the premise of obtaining the probability distribution map of the reference image, the illumination estimation calculation of the captured image and the gain calculation of each channel of the captured image are very simple and fast, and can meet real-time calculations to adjust Capture image white balance needs.

Embodiment four

As shown in FIG. 9, corresponding to the white balance method of the above-mentioned embodiment, an embodiment of the present invention provides a white balance device, including:

The gray point probability value acquisition unit 91 is configured to obtain the gray point probability value of each pixel in the captured image from the gray point probability distribution map of the reference image according to the color values of each pixel in the captured image.

The illumination estimation unit 92 is configured to obtain an illumination estimate of the captured image according to the gray point probability value of each pixel in the captured image.

The gain value obtaining unit 93 is configured to obtain the gain value of each color channel of the captured image according to the illumination estimation of the captured image, so as to adjust the captured image to achieve white balance.

Optionally, the gray point probability distribution map of the reference image can be obtained by the device described in the second embodiment.

Moreover, in the gray-scale point probability value acquisition unit 91, the shooting scene mode for obtaining the captured image is consistent with the preset shooting scene mode for obtaining the gray-scale point probability distribution map of the reference image. If the lighting is all outdoor lighting or indoor lighting, no repeat.

Optionally, as shown in FIG. 10, in the white balance device of the embodiment of the present invention, the gray point probability value acquisition unit 91 may include:

The first acquiring subunit 911 is configured to determine the corresponding position of each pixel in the gray point probability distribution map of the reference image according to each color value of each pixel in the captured image.

The second obtaining subunit 912 is configured to obtain the gray point probability value corresponding to each pixel point according to the corresponding position of each pixel point in the gray point probability distribution map of the reference image.

Further, the illumination estimating unit 92 is specifically used to calculate the weighted average value of each color value of all pixels by using the gray point probability value of each pixel in the captured image as a weight, and obtain the average value of each color channel of the captured image as the captured image. Lighting estimation for images.

Optionally, the gain value obtaining unit 93 may include:

The gain value acquisition subunit 931 is configured to take the reciprocal of the average value of each color channel of the captured image to obtain the gain value of each color channel.

The gain control subunit 932 is configured to control the gain value of each color channel obtained by the gain value acquisition subunit 931 to adjust the captured image to achieve white balance.

The white balance device of the embodiment of the present invention may also include:

The gray point probability distribution map storage unit 94 is configured to store the gray point probability distribution map of the reference image.

The image output unit 95 outputs the image for the next step of processing after the captured image is white-balanced.

The white balance device of the embodiment of the present invention may also include:

The image acquisition unit 96 may be a CCD (Charge Couple Device, charge-coupled device) or a CMOS (Complementary Metal-Oxide Semiconductor, complementary metal oxide semiconductor) photosensitive device. The image acquisition unit outputs digital image information to the first acquisition subunit 911 of the gray point probability value acquisition unit 91 .

It can be known that if the white balance device of the embodiment of the present invention is used together with the device for obtaining the probability distribution of image gray points in Embodiment 2, the image acquisition unit 96 of the white balance device of the embodiment of the present invention can be used together with the device for obtaining the image gray point probability distribution. The image acquisition unit 61 of the device for point probability distribution is the same functional unit, which will not be repeated here.

Specifically, the white balance device provided by the embodiments of the present invention:

The image acquisition unit 96 obtains digital image information (hereinafter referred to as the captured image), and sends the image to the first acquisition subunit 911 of the gray point probability value acquisition unit 91;

The first acquisition subunit 911 determines the gray point of each pixel in the reference image according to the color values of each pixel in the captured image and the gray point probability distribution map of the reference image in the gray point probability distribution map storage unit 94. The corresponding position in the probability distribution map, and send the position information to the first acquisition subunit 912;

The first acquisition subunit 912 obtains the gray point probability value corresponding to each pixel point according to the corresponding position of each pixel point in the gray point probability distribution map of the reference image, and sends the gray point probability value to the color channel value acquisition Unit 92;

The illumination estimating unit 92 calculates the weighted average value of each color value of all pixels with the gray point probability value of each pixel point as the weight, obtains the average value of each color channel of the captured image, and sends the average value of each color channel of the captured image to the gain The gain value acquisition subunit 931 of the value acquisition unit 93;

The gain value acquisition subunit 931 takes the reciprocal of the average value of each color channel to obtain the gain value of each color channel, and sends the gain value of each color channel to the gain control subunit 932;

The gain control subunit 932 controls and adjusts the captured image to achieve white balance according to the gain value of each color channel, and sends the white-balanced captured image to the image output unit 95;

After the image is captured to achieve white balance, subsequent steps can be performed, which will not be repeated here.

It can be seen from the white balance device provided by the above-mentioned embodiments of the present invention that the gray point probability value of each pixel in the image can be obtained according to the gray point probability distribution map of the reference image, so that the judgment of the gray point is more accurate , thereby improving the accuracy of white balance gain calculation.

Moreover, in the white balance device provided by the embodiments of the present invention, on the premise of obtaining the probability distribution map of the reference image, the gain calculation of each color channel of the image is very simple and fast, and can meet the needs of real-time calculation.

Embodiment five

Referring to Fig. 1-5, Fig. 7, and Fig. 8, the white balance method provided by the embodiment of the present invention is described by taking the R, G, and B channel values of the RGB chromaticity space as an example in which the illumination is estimated.

As shown in Figure 11, first, the method for obtaining the probability distribution map of the gray points of the reference image is described:

Step 001, photograph the standard 24 color palettes to obtain a reference image.

It should be understood that, considering the pursuit of different levels of adjustment effects, or for the convenience of operation, the reference image can also be obtained by shooting other objects or images with color blocks and grayscale blocks, such as in the shooting place Fixed reference objects and other objects that are easy to carry or photograph with color blocks and grayscale blocks.

Step 002. Obtain the grayscale color block and color block histogram of the RGB chromaticity space reference image respectively.

For the gray-scale color block and color block of the reference image of the standard 24-color plate, the following operations are performed respectively: For each pixel in the color block, two values of R÷G and B÷G are calculated, and R÷G is The first dimension, B÷G is the second dimension, and calculate the two-dimensional histogram of all the pixels in the color block, as shown in Figures 2 and 3.

Step 003. Obtain the number of pixels in each interval of the color point and gray point histogram.

Referring to Figures 2 and 3, the value range of each coordinate axis in the histogram is [0-2.5], and each interval is Delta=0.005, then the histogram has a total of 501×501 intervals (each coordinate axis of the histogram The maximum value is 500).

Referring to the value of interval A in the histogram in FIG. 4 , it represents the number of all pixels falling into (2*delta)<R÷G<(3*delta), delta<B÷G<(2*delta).

It is also possible to normalize each histogram so that the sum of all points in the histogram is 1, reducing the amount of calculation, that is, calculating for a certain point x to get

Step 004. Obtain a probability distribution map of gray points of each pixel in the reference image.

If the value of this position in the grayscale histogram is a, and the value of this position in the color point histogram is b, then the probability value of this position as a grayscale point is p=a÷(a+b).

When the case of (a+b)=0 occurs, it is considered that p=0. When the pixel points outside the histogram expression range, that is, beyond the range specified by [0, 2.5], p=0 is also taken.

Finally, the gray point probability distribution map of the reference image is obtained.

It is also possible to filter the reference gray point probability distribution map to make the reference probability distribution map smoother. Alternatively, the reference gray point probability distribution map may be stored, and the reference gray point probability distribution map may be called when white balance adjustment is performed.

Next, explain the method of white balance:

Step 005. Obtain the gray point probability value of each pixel of the input image from the gray point probability distribution map of the reference image.

The input image is an image to be adjusted for white balance, and the input image may be a captured image. The captured image is taken as an example below. Calculate the R÷G value and B÷G value of the pixel point of the captured image, and find the corresponding gray point probability value p from the reference gray point probability distribution map according to the R÷G value and B÷G value.

Step 006. Obtain an illumination estimation result of the captured image.

P_gray(C_i)为像素点的灰度点概率值，C_i，k为像素点各颜色值，k为R或G或B，i为像素点个数编号。拍摄图像的光照估计为： $L_R=\frac{\underset{i}{\mathrm{\&Sigma;}}{P}_{\mathrm{gray}}\left(C_i\right)*C_{i,R}}{\underset{i}{\mathrm{\&Sigma;}}1},$ $L_G=\frac{\underset{i}{\mathrm{\&Sigma;}}{P}_{\mathrm{gray}}\left(C_i\right)*C_{i,G}}{\underset{i}{\mathrm{\&Sigma;}}1},$ $L_B=\frac{\underset{i}{\mathrm{\&Sigma;}}{P}_{\mathrm{gray}}\left(C_i\right)*C_{i,B}}{\underset{i}{\mathrm{\&Sigma;}}1}.$ The calculation formula for the illumination estimation of the captured image is

P _gray (C _i ) is the probability value of the gray point of the pixel, C _{i, k} is the color value of the pixel, k is R or G or B, and i is the number of the pixel. The illumination estimate for the captured image is: $L_R=\frac{\underset{i}{\mathrm{\&Sigma;}}{P}_{\mathrm{gray}}\left(C_i\right)*C_{i,R}}{\underset{i}{\mathrm{\&Sigma;}}1},$ $L_G=\frac{\underset{i}{\mathrm{\&Sigma;}}{P}_{\mathrm{gray}}\left(C_i\right)*C_{i,G}}{\underset{i}{\mathrm{\&Sigma;}}1},$ $L_B=\frac{\underset{i}{\mathrm{\&Sigma;}}{P}_{\mathrm{gray}}\left(C_i\right)*C_{i,B}}{\underset{i}{\mathrm{\&Sigma;}}1}.$

Step 007, obtain the gain values of the R, G, and B channels of the captured image.

For example, the gain value of the R channel, R _gain =1/L _R .

For example, the gain value G _gain of the G channel =1/L _G .

For example, the gain value B _gain of the B channel =1/L _B .

By adjusting the captured image according to the gain values of the R, G, and B channels, an image with accurate white balance can be obtained.

It can be known that under the premise of obtaining the probability distribution map of the reference image, the calculation of the gain of each color channel of the captured image is very simple and fast, and can meet the needs of real-time calculation.

Step 008, outputting the captured image after white balance adjustment.

Those of ordinary skill in the art can understand that the histogram established in the above steps can adopt different color coordinate systems, such as (R-G) ÷ G in RGB chromaticity space, (B-G) ÷ G coordinates, or Cb, Cr in Ycbcr space coordinates, do not repeat them.

Embodiment six

The difference between this embodiment and the fifth embodiment is that in the method of obtaining the probability distribution map of image gray points in this embodiment, the reference image probability distribution map of the standard 24-color palette can be obtained under outdoor lighting or indoor lighting conditions respectively , and can respectively store the reference image probability distribution map of the standard 24 color palettes for outdoor lighting or indoor lighting.

When performing photography operations, the user can select the current shooting scene mode as outdoor lighting or indoor lighting, so that the automatic white balance of the image can be realized according to the gray point probability distribution map of the selected scene mode.

Embodiment seven

The difference between this embodiment and Embodiment 5 and Embodiment 6 is that in the method of obtaining the probability distribution map of image gray points in this embodiment, sunlight on sunny days, sunlight or shadows on cloudy days, tungsten lighting, and daylight lighting can be respectively obtained. Or fluorescent lighting, etc., the reference image probability distribution map of the standard 24 color plates, and can store the probability distribution map of each reference image separately.

In this way, the current shooting scene mode that the user can select is more detailed and clear.

When performing photography operations, the user can select the current shooting scene mode as sunny daylight, cloudy daylight or shadow, tungsten light, daylight light or fluorescent light, etc., so that the probability distribution of gray points according to the selected scene mode Figure, to achieve automatic image white balance.

In summary, from the technical solutions provided by the embodiments of the present invention, it can be seen that since the gray point probability distribution map of the reference image of the standard color palette is obtained, the gray point probability value of each pixel in the captured image It can be obtained according to the gray point probability distribution map of the reference image, so that the judgment of the gray point is more accurate, and the accuracy of white balance gain calculation is improved.

It is easy to understand that the various histograms and distribution diagrams described in the embodiments of the present invention include charts, plots, numbers and other forms that can be used to represent the distribution of numerical values in a region.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented through computer programs to instruct related hardware, and the programs can be stored in a computer-readable storage medium. During execution, it may include the processes of the embodiments of the above-mentioned methods. Wherein, the storage medium may be a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM) or a random access memory (Random Access Memory, RAM), etc.

The above is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present invention can easily think of changes or Replacement should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be determined by the protection scope of the claims.

Claims (21)

1. A method for obtaining the probability distribution of image gray points, characterized in that, comprising: Take a reference image with color patches and grayscale patches; Establish a color point histogram according to the color values of each pixel corresponding to the color block in the reference image, and establish a gray point according to the color values of each pixel corresponding to the gray color block in the reference image histogram; Dividing the color point histogram and the gray point histogram into a plurality of intervals respectively, according to the number of pixels in the corresponding intervals in the color point histogram and the gray point histogram, A gray point probability distribution diagram of each pixel point in the reference image is obtained. 2. The method for obtaining image gray point probability distribution according to claim 1, wherein the two dimensions of the color point histogram and the gray point histogram include: R÷G and B÷G, or (R-G)÷G and (B-G)÷G of each pixel in the RGB chromaticity space; Alternatively, each pixel is in Cb and Cr in the YCbCr chromaticity space. 3. The method for obtaining the probability distribution of image grayscale points according to claim 1 or 2, characterized in that, according to the number of pixels in the corresponding intervals in the color point histogram and the grayscale point histogram , to obtain the gray point probability distribution map of the reference image, including: The number of pixels in the interval in the grayscale point histogram is a, and the number of pixels in the corresponding interval in the color point histogram is b, then the pixels in the interval are the probability values of grayscale points It is p=a÷(a+b). 4. The method for obtaining the probability distribution of image gray points according to claim 1 or 2, wherein the reference image is obtained by photographing a standard color palette with color patches and grayscale patches. 5. The method for obtaining the probability distribution of image grayscale points according to claim 4, wherein said photographing a reference image with color patches and grayscale patches comprises: under preset illumination, photographing a reference image with color patches The standard color palette of the color decision and the grayscale color block is used to obtain a reference image. The preset lighting includes: outdoor lighting or indoor lighting, the outdoor lighting includes: sunny daylight, cloudy daylight or shadow, and the indoor lighting includes: Tungsten lighting, daylight lighting or fluorescent lighting. 6. A white balance method, characterized in that, comprising: According to each color value of each pixel point in the captured image, and the gray-scale point probability distribution map of the reference image, the gray-scale point probability value of each pixel point in the captured image is obtained; Obtaining an illumination estimate of the captured image according to the gray point probability value of each pixel in the captured image; According to the illumination estimation of the captured image, the gain value of each color channel of the captured image is obtained to adjust the captured image to achieve white balance. 7. The white balance method according to claim 6, characterized in that, according to each color value of each pixel in the captured image and the gray point probability distribution map of the reference image, the value of each pixel in the captured image is obtained. Gray point probability value, including: According to each color value of each pixel point in the captured image, determine the corresponding position of each pixel point in the gray point probability distribution map of the reference image; According to the corresponding position of each pixel point in the gray point probability distribution map of the reference image, the gray point probability value corresponding to each pixel point is obtained. 8. The white balance method according to claim 6, wherein the illumination estimation of the captured image is obtained according to the gray point probability value of each pixel in the captured image, comprising: Using the gray point probability value of each pixel in the captured image as a weight, calculate the weighted average of each color value of all pixels, and obtain the average value of each color channel of the captured image as the illumination estimate of the captured image. 9. The white balance method according to claim 6, wherein the gain value of each color channel of the captured image is obtained according to the illumination estimation of the captured image to adjust the captured image to achieve white balance, comprising: Taking the inverse of the average value of each color channel of the captured image to obtain the gain value of each color channel; The captured image is controlled by adjusting the gain value of each color channel to achieve white balance. 10. The white balance method according to any one of claims 6-9, wherein the color channels of the image include the R channel, the G channel and the B channel of the RGB chromaticity space, or the YCbCr chromaticity space Cb channel and Cr channel. 11. A white balance device, comprising: A gray-scale point probability value acquisition unit, configured to obtain the gray-scale point probability value of each pixel in the captured image according to the color values of each pixel in the captured image and the gray-scale point probability distribution map of the reference image; An illumination estimation unit, configured to obtain an illumination estimation of the captured image according to the gray point probability value of each pixel in the captured image; The gain value obtaining unit is configured to obtain the gain value of each color channel of the captured image according to the illumination estimation of the captured image, so as to adjust the image to achieve white balance. 12. The white balance device according to claim 11, wherein the gray point probability value acquisition unit comprises: The first acquisition subunit is configured to determine the corresponding position of each pixel in the gray point probability distribution map of the reference image according to each color value of each pixel in the captured image; The second obtaining subunit is configured to obtain the gray point probability value corresponding to each pixel point according to the corresponding position of each pixel point in the gray point probability distribution map of the reference image. 13. The white balance device according to claim 11, characterized in that the illumination estimation unit is specifically configured to use the gray point probability value of each pixel in the captured image as a weight to calculate the weighted average of each color value of all pixels value, and the average value of each color channel of the captured image is obtained as the illumination estimation of the image. 14. The white balance device according to claim 11, wherein the gain value acquisition unit comprises: The gain value acquisition subunit is used to take the reciprocal of the average value of each color channel of the captured image to obtain the gain value of each color channel; The gain control subunit is used to adjust the gain value of each color channel obtained by the gain value acquisition subunit to control the captured image to achieve white balance. 15. The white balance device according to any one of claims 11-14, wherein the color channels of the image include the R channel, the G channel and the B channel of the RGB chromaticity space, or the YCbCr chromaticity space Cb channel and Cr channel. 16. The device for obtaining the probability distribution of image grayscale points according to claim 11, wherein the grayscale point probability value acquisition unit comprises an image acquisition unit for photographing images with color patches and grayscale patches. Reference image: a histogram building unit, used to create a color point histogram according to the color values of each pixel corresponding to the color color block in the reference image, and according to each color point corresponding to the gray color block in the reference image The color value of the pixel point is used to establish a gray point histogram; the histogram analysis unit is used to divide the color point histogram and the gray point histogram into multiple intervals respectively, and divide the pixels in the interval The number of points is used as the value of the interval; the gray point probability calculation unit is used to obtain the gray point probability distribution map of the reference image according to the value of the corresponding interval of the color point histogram and the gray point histogram . 17. The device for obtaining the probability distribution of image gray points according to claim 16, wherein the color point histogram established by the histogram building unit and the two dimensions of the gray point histogram include: R÷G and B÷G, or (R-G)÷G and (B-G)÷G of each pixel in the RGB chromaticity space; Alternatively, each pixel is in Cb and Cr in the YCbCr chromaticity space. 18. The device for obtaining the probability distribution of image gray-scale points according to claim 16, wherein the gray-scale point probability calculation unit is used for the number of pixels in the interval of the gray-scale point histogram to be a, the number of pixels in the corresponding interval in the color point histogram is b, and the calculated probability value of the pixel in the interval being a grayscale point is p=a÷(a+b). 19. A device for obtaining the probability distribution of image gray points, comprising: An image acquisition unit, used for taking reference images with color patches and grayscale patches; A histogram building unit, configured to create a color point histogram according to the color values of the pixels corresponding to the color blocks in the reference image, and to establish a histogram of color points according to the color values of the pixels corresponding to the gray color blocks in the reference image. Color value, build gray point histogram; a histogram analysis unit, configured to equally divide the color point histogram and the gray point histogram into multiple intervals, and obtain corresponding intervals in the color point histogram and the gray point histogram The number of pixels in The gray point probability calculation unit is configured to obtain the gray point probability distribution map of the reference image according to the number of pixels in the color point histogram and the corresponding interval in the gray point histogram. 20. The device for obtaining the probability distribution of image gray points according to claim 19, wherein the color point histogram established by the histogram building unit and the two dimensions of the gray point histogram include: R÷G and B÷G, or (R-G)÷G and (B-G)÷G of each pixel in the RGB chromaticity space; Alternatively, each pixel is in Cb and Cr in the YCbCr chromaticity space. 21. The device for obtaining the probability distribution of image gray-scale points according to claim 19, wherein the gray-scale point probability calculation unit is used for the number of pixels in the interval of the gray-scale point histogram to be a, the number of pixels in the corresponding interval in the color point histogram is b, and the calculated probability value of the pixel in the interval being a gray point is p=a÷(a+b).

Images