WO2019062633A1 - Color shading correction method and device - Google Patents

Abstract

Provided in the present application are a color shading correction method and device, capable of accurately correcting color shading of an image. The method comprises: receiving a first image; extracting a hue value of a first pixel point in the first image for which a gradient is less than or equal to a threshold; determining a transformation coefficient of the first pixel point by using the hue value of the first pixel point and a feature contour of the first pixel, wherein the feature contour is used for representing a feature of a color shading contour of a pixel point in the image that is obtained by statistically analyzing multiple standard light sources, and the transformation coefficient is used for representing the degree of strength of the feature contour; determining a color shading contour of a second pixel point in the first image according to the transformation coefficient of the first pixel point and a color shading model; correcting a color shading of the second pixel point according to the color shading contour of the second pixel point.

Description

Method and device for color shading correction

The present application claims priority to Chinese Patent Application No. 200910911423.2, filed on Sep. 29, 2017, the entire disclosure of which is incorporated herein by reference. .

Technical field

The present application relates to the field of image processing and, more particularly, to methods and apparatus for color shading correction in the field of image processing.

Background technique

Due to the thin and light design requirements of wireless terminals and other mobile devices, the size of the digital camera module is required to be smaller and smaller, which brings various side effects to the output image, one of which is color shading. Color shading refers to the color gradient from the center to the edge in the image output by the color camera module. Generally, in the camera module, the front side of the digital image sensor is equipped with an infrared cut-off (IR-CUT) filter, which blocks the infrared light while transmitting visible light to ensure the accuracy of color reproduction. However, the cutoff wavelength of the infrared cut filter varies with the angle of the incident light, causing the sensor center to be different from the edge cutoff wavelength, which is represented by the color shading of the image. Moreover, the reduction in the size of the camera module increases the angle of the incident light, making the color shadow more prominent. In addition, mismatching of devices such as the main lens and microlenses can also add color shading.

The degree of color shading changes as the reflectance spectrum of the scene changes, so the color shading needs to be dynamically corrected. Currently, the prior art corrects color shading based on a color consistency model that includes a set of general contours and a set of transform coefficients corresponding to color temperatures. However, the color temperature and the spectrum are not one-to-one correspondence. For example, the A-light and U30 color temperatures are close but the spectral difference is large, which causes the color consistency model to be rough, resulting in inaccurate color shading correction using the model.

Summary of the invention

The present application provides a method and apparatus for color shading correction that accurately corrects color shading of an image.

In a first aspect, a method of color shading correction is provided, the method comprising:

Receiving a first image;

Extracting a tone value of the first pixel in the first image, wherein a gradient of a tone value of the first pixel is less than or equal to a threshold;

Determining a transform coefficient of the first pixel point by using a tone value of the first pixel point and a feature profile of the first pixel point, wherein the feature profile is used to represent an image of a plurality of statistically derived standard light sources a feature of a color shaded outline of a pixel in which the transform coefficient is used to indicate a degree of strength of the feature outline;

Determining a color shadow profile of a second pixel point in the first image according to a transform coefficient and a color shadow model of the first pixel, wherein the color shadow model is used to represent a color shadow of a pixel point in the image a mapping relationship between contours, feature contours, and transform coefficients;

Correcting the color shading of the second pixel point according to the color shading profile of the second pixel point.

The embodiment of the present application first extracts a tone value of the first pixel of the first image, and then determines a model parameter of the first pixel according to the tone value of the first pixel and the feature contour of the first pixel (ie, transform a coefficient), further the model parameter and the color shading model of the first pixel, determining a color shading contour of the first image, and finally correcting the color shading of the first image according to the color shading contour of the first image. Since the embodiment of the present application determines the model parameters by directly using the tonal value of the image, and does not rely on other modules such as white balance to acquire the model parameters, the embodiment of the present application can accurately determine the model parameters. Moreover, in the embodiment of the present application, the feature contour can represent the color shadow caused by a specific factor or a statistical feature, and the color shadow model can accurately represent the complex color shadow contour, and thus the embodiment of the present application can accurately image the color shadow of the image. Make corrections.

Optionally, before determining the transform coefficient of the first pixel by using the tone value of the first pixel and the feature contour of the first pixel, the method further includes:

Extracting a color shadow outline of a pixel point of the flat field image of the plurality of standard light sources;

Determining a feature vector of a color shadow profile of a pixel of the flat field image of the plurality of standard light sources, and determining the feature profile based on the feature vector.

Thus, the feature contour can be used to represent the characteristics of the color shaded contours of the images of the plurality of standard light sources obtained by statistics. That is, each feature profile is a feature that is statistically derived from the color shadow profile of an image of multiple standard sources. At this time, each feature contour may correspond to a color shadow caused by a specific physical factor or a statistical feature, wherein the physical factor is, for example, a change in incident angle, a mismatch of a device such as a main lens and a microlens, and the like.

Optionally, determining, by using a tone value of the first pixel and a feature profile of the first pixel, a transform coefficient of the first pixel, including:

Determining the transform coefficients of the first pixel according to the following formula:

为x的变换系数且

f _k(x)为x的第k维的特征轮廓，a _k为f _k(x)的变换系数，b _k为f _k(x)的平移系数，且a _k+b _k＝1，k的取值范围为1至N且代表所述特征轮廓的维度，N为大于1的正整数，H(x)表示x处的色调值的对数。 Where x represents the first pixel point,

Is the transform coefficient of x and

F _k (x) is the k-dimensional feature profile of x, A _k is a transform coefficient F _k (x) is, B _k is F _k (x) the translation factor and _{a k + b k = 1,} k is The value ranges from 1 to N and represents the dimension of the feature profile, N is a positive integer greater than 1, and H(x) represents the logarithm of the tone value at x.

则有： Specifically, for the rg hue, the transform coefficient of the first region can be defined.

Then there are:

则有： For bg hue, the transform coefficient of the first region can be defined

Then there are:

Optionally, in the embodiment of the present application, the transform coefficient of the first pixel in the first region may be used as the transform coefficient of the second pixel in the acquired first image. That is to say, in the embodiment of the present application, the transform coefficients of the respective pixel points in the image are the same. At this time, the transform coefficient and the feature contour of the second pixel point can be brought into the image color shadow model to synthesize the color shadow outline of the image.

Specifically, the color shading model can have the following form:

Wherein, s (x) denotes a color shading profile of x, f _k (x) for the first K x dimension of the feature profile, a _k is f _k (x) of transform coefficients, b _k is f _k (x) is translated The coefficient, k, ranges from 1 to N and represents the number of features, and N is a positive integer greater than one. In the embodiment of the present application, it is agreed that the colorless shadow f _k (x)=1, in order to maintain the normalized property of f _k (x), a _k +b _k =1.

In the embodiment of the present application, the transform coefficient a _{k is} used to describe the degree of strength of the feature contour f _k (x) corresponding to the transform coefficient a _k . Therefore, the color shading model in the embodiment of the present application can characterize a plurality of factors or color shading components when statistical features are superimposed by multiplying a plurality of feature contours.

At this time, the color shadow outline of the second pixel point is:

Wherein, x 'representing the second pixel, s (x') represented by x 'shadow color profile, f _k (x') as x 'of the k-dimensional feature profile, a _k is F _k (x' The transform coefficient, b _k is the translation coefficient of f _k (x'), and a _k +b _k =1, k ranges from 1 to N and represents the number of features, N is greater than 1 Positive integer.

Optionally, the correcting the color shading of the second pixel point according to the color shadow profile of the second pixel point comprises:

A reciprocal of a color shading profile of the second pixel point is determined as a correction value of the second pixel point. Specifically, the correction table may include an image pixel point and a correction value of the pixel point, and the correction value of the pixel point may specifically be a reciprocal of the color shadow profile of the pixel point.

Correcting the color shading of the second pixel point according to the correction value of the second pixel point. Specifically, the product of the value of the R channel of the second pixel point x′ and the correction value of the rg tone is the value of the R channel of the second pixel point x′ after the correction, and the value of the B channel of the second pixel point x′ and bg The product of the correction value of the hue is the value of the B channel of the second pixel point x' after correction.

In a second aspect, the embodiment of the present application provides a device for performing color shading correction, which is used to perform the method in any of the foregoing first aspect or any possible implementation manner of the first aspect, and specifically, the device includes A module of a method in any aspect or in any possible implementation of the first aspect.

In a third aspect, an embodiment of the present application provides a color shading correction apparatus, including: a memory and a processor. Wherein the memory is for storing instructions for executing the instructions stored by the memory, and when the processor executes the instructions stored by the memory, the executing causes the processor to perform any of the first aspect or the first aspect The method in the implementation.

In a fourth aspect, the embodiment of the present application provides a computer readable medium for storing a computer program, the computer program comprising instructions for executing the method in the first aspect or any possible implementation manner of the first aspect.

DRAWINGS

FIG. 1 is a schematic diagram of an application scenario of color shading correction according to an embodiment of the present application.

2 is a schematic flow chart of a method for calibrating a color shading model according to an embodiment of the present application.

FIG. 3 is a schematic flowchart of a method for color shading correction according to an embodiment of the present application.

4 is a schematic block diagram of an apparatus for color shading correction in an embodiment of the present application.

FIG. 5 is a schematic block diagram of another apparatus for color shading correction according to an embodiment of the present application.

Detailed ways

The technical solutions in the present application will be described below with reference to the accompanying drawings.

FIG. 1 is a schematic diagram of an application scenario of color shading correction according to an embodiment of the present application. Specifically, the light reflected by the scene 11 is projected through the lens 12, and a digital image signal is formed on the image sensor 13, and the digital image signal is first subjected to a dead pixel correction, black level compensation, etc. through the preprocessing module 14, and then the preprocessing is performed. The image signal is input to a color shading estimation module 15, which estimates the color shading of the image signal to obtain a correction table for the image. The correction module 16 corrects the image signal based on the correction table determined by the color shading estimation module. The subsequent processing module 17 further performs subsequent processing on the corrected image signal to generate a final image 18.

The color shadow estimation module 15 has been improved in the embodiment of the present application. The color shadow estimation module 15 can be applied to an Image Signal Processing (ISP) chip for processing image signals collected by the color camera module. The image signal processing chip can be installed separately or integrated into a mobile phone, digital camera, tablet, personal digital assistant or other device that can capture images, or can be installed separately or integrated into a desktop computer, video conferencing station or other internal or external camera. In the device to capture images. Therefore, the apparatus having the image processing chip can perform color shading correction on a still image or a video image.

In general, the way color shading works can be expressed by the following formula (1)

h(x)=s(x)·i(x) (1)

Where x is the pixel point, i(x) is the ideal image signal of x without color shading, s(x) is the color shadow outline of x, which may also be called color shading component or color shading information, h(x) is The actual image signal of x. Also, in the formula (1), the image signal may specifically be a tone value of the image.

In the embodiment of the present application, the color shading estimation module 15 may correct the color shading of the image based on the color shading model. Here, the color shading model can be used to represent the mapping relationship of the color shading outline, the feature outline, and the transform coefficients of the image. Specifically, the color shading model can have the following form:

Wherein, s (x) denotes a color shading profile of x, f _k (x) for the first K x dimension of the feature profile, a _k is f _k (x) of transform coefficients, b _k is f _k (x) is translated The coefficient, k, ranges from 1 to N and represents the number of features, and N is a positive integer greater than one. In the embodiment of the present application, it is agreed that the colorless shadow f _k (x)=1, in order to maintain the normalized property of f _k (x), a _k +b _k =1.

In particular, the feature contours in equation (2) can be used to represent features of the color shaded contours of the images of the plurality of standard light sources obtained by statistics. That is, each feature profile is a feature that is statistically derived from the color shadow profile of an image of multiple standard sources. At this time, each feature contour may correspond to a color shadow caused by a specific physical factor or a statistical feature, wherein the physical factor is, for example, a change in incident angle, a mismatch of a device such as a main lens and a microlens, and the like. And, the transform coefficient a _{k is} used to describe the degree of strength of the feature contour f _k (x) corresponding to the transform coefficient a _k . Therefore, the color shading model in the embodiment of the present application can characterize a plurality of factors or color shading components when statistical features are superimposed by multiplying a plurality of feature contours.

It should be noted that the color shading model in the embodiment of the present application is not limited to the form of the above formula (2), and may be, for example, various variants of the formula (2) (for example, logarithm of f _k (x) or s(x) The embodiment of the present application does not limit this.

FIG. 2 is a schematic flow chart showing a method for calibrating a feature contour in a color shading model according to an embodiment of the present application. It should be understood that FIG. 2 illustrates the steps or operations of the method of calibrating the feature contours in the color shading model, but these steps or operations are merely examples, and other embodiments of the present application may also perform other operations or variants of the operations in FIG. . Moreover, the various steps in FIG. 2 may be performed in a different order than that presented in FIG. 2, and it is possible that not all operations in FIG. 2 are to be performed.

201, collecting an image.

Specifically, the image sensor 13 of FIG. 1 can be used to acquire an image. Specifically, a standard light source is projected onto the scene 11, and the scene 11 projects light of the standard light source through the lens 12 to form an image of the standard light source on the image sensor 13. Since the cutoff wavelength of the infrared cut filter on the image sensor 13 becomes shorter as the incident angle increases, and the reduction in the size of the camera module increases the incident light angle, the image acquired in 201 has a color shading.

Here, a flat field image of each standard light source can be acquired. That is to say, the scene 11 at this time may be a scene with no color and a flat surface such as white paper or white wall. Thus, the actual color at different locations in the image is related to the color shading component of the location, and the actual color is independent of the properties of the scene 11 itself (eg, surface color and curvature of the surface, etc.).

Here, the standard light source is an artificial light source that simulates various ambient light, such as American kitchen window spotlight A, family hotel lamp F, simulated sunlight D50, and international standard artificial daylight D65, which simulates the average northern sunlight. D75, European Standard Warm White TL83, European, Japanese, Chinese store light source TL84, American Cool White Fluorescent (CWF), American Warm White Fluorescent (U30), USA Retailer Target - Target specifies light source U35, UV-light source (Ultra-Violet, UV) and other light sources. Here, a flat field image of each standard light source can be acquired. It can be understood that different standard light sources have different spectra, color temperatures, and power.

In the embodiment of the present application, at least two standard light sources can be simultaneously projected on the scene, so that a flat field image in which at least two standard light sources work together can be formed on the image sensor. In the embodiment of the present application, one or more flat field images may be acquired for one standard light source, or one or more flat field images may be acquired when at least two standard light sources work together.

202. Extract a color shadow outline of the image.

Specifically, the color shadow outline of each pixel in the image can be extracted. First, the Red Green Blue (RGB) value of each pixel in the image acquired in 201 can be obtained. When a plurality of flat field images of a standard light source or at least two standard light sources are combined, the RGB average value of each pixel point can be obtained for the plurality of flat field images, and the RGB average of each pixel point is obtained. The value is the RGB value of each pixel of the flat field image when the standard source or at least two standard sources are combined. Thus, for a standard source, a flat field image can be obtained. When at least two standard light sources work together, a corresponding flat field image can be obtained.

Here, the RGB value of all the pixels in the image can be referred to as the RGB value of the image. After the RGB values of each flat field image are acquired, the RGB values can be converted to tone values. Specifically, the rg tone component and the bg tone component of the pixel point can be obtained according to the RGB value of one pixel point, the rg tone component is a ratio of the R value to the G value, and the bg tone component is a ratio of the B value to the G value, that is, Rg=R/G, bg=B/G.

其中j＝1,…,M。对于每个像素点的bg色调，可以获取该像素点的颜色阴影轮廓

其中j＝1,…,M。 For each color point of each pixel in the flat field image, normalization processing can be performed on a basis of each optical center, respectively, to obtain a color shadow outline of each pixel. Here, the number of flat field images may be M, and M is a positive integer greater than one. Specifically, for the rg tone of each pixel, the color shadow outline of the pixel can be obtained.

Where j=1,...,M. For the bg hue of each pixel, you can get the color shadow outline of the pixel

Where j=1,...,M.

203. Determine whether the color shadow outlines of all standard light sources are extracted.

If not, repeat steps 201 and 202 above for other standard sources. If yes, proceed to the next step 204.

204. Extract feature contours. Specifically, the feature contour of each of the above pixel points can be extracted.

将所有标准光源平场图像的rg色调的颜色阴影轮廓

j＝1,2，…,M视为一个集合，即平场图像的rg色调的颜色阴影轮廓集合可以表示为矩阵

Specifically, the logarithm of the color shading profile of the rg tone of each standard light source flat field image extracted can be obtained.

Color shaded outline of rg tones of all standard light source flat field images

j=1, 2,..., M is regarded as a set, that is, the color shadow outline set of the rg tone of the flat field image can be expressed as a matrix

的特征向量

其中N ₁表示该特征向量的维度，特征向量的维度可以理解为该特征向量的个数，且为小于M的正整数。本申请实施例中，可以使用主成分分析(Principal Component Analysis，PCA)、卷积神经网络(Convolutional Neural Networks，CNN)等比较公知的方法求该特征向量，具体的，求取特征向量的过程可以参见现有技术中的求矩阵的特征向量的方法，本申请实施例不再赘述。 Then, find the matrix

Characteristic vector

Where N ₁ represents the dimension of the feature vector, and the dimension of the feature vector can be understood as the number of the feature vector and is a positive integer smaller than M. In the embodiment of the present application, the feature vector can be obtained by using a well-known method such as Principal Component Analysis (PCA) or Convolutional Neural Networks (CNN). Specifically, the process of obtaining the feature vector can be performed. For the method of the eigenvectors of the matrix in the prior art, the embodiments of the present application are not described again.

的指数，得到图像的rg色调的特征轮廓

这里N ₁表示rg色调的特征轮廓的维度或个数。 Finally, you can find the eigenvector

The index that gives the characteristic outline of the rg tones of the image

Here N ₁ represents the dimension or number of feature contours of the rg hue.

其中N ₂为小于M的正整数，表示bg色调的特征轮廓的维度或个数。 In the same way, the characteristic contour of the bg tone of the image can be obtained in the same way as the rg hue.

Where N ₂ is a positive integer less than M, representing the dimension or number of feature contours of the bg hue.

Here, the description of the feature contour can be referred to the description above, and to avoid repetition, it will not be repeated here.

205. Store the feature contours of the above rg hue and bg hue. Specifically, the feature contour of each tone of each pixel point can be stored.

Therefore, the embodiment of the present application extracts a flat field image of a plurality of standard light sources, then extracts a color shadow profile of the flat field image of each standard light source, and extracts a set of feature contours according to the set of color shadow contours of all standard light sources. Feature contours can characterize color shadows caused by a particular factor or statistical feature.

Further, the above set of feature contours and a set of transform coefficients are combined by multiplication to form a color shading model described in the embodiment of the present application. In the embodiment of the present application, the transform coefficient can represent the strength of the factor or the statistical feature, and the multiplication can reflect the physical rule when the different factors are superimposed. Therefore, the color shadow model in the embodiment of the present application can accurately represent the complex color shadow. profile.

FIG. 3 is a schematic flowchart of a method for color shading correction according to an embodiment of the present application. Specifically, the method can be performed by the color shading estimation module 15 and the correction module 16 shown in FIG. It should be understood that FIG. 3 illustrates the steps or operations of the method of color shading correction, but these steps or operations are merely examples, and other embodiments of the present application may also perform other operations or variations of the various operations in FIG. Moreover, the various steps in FIG. 3 may be performed in a different order than that presented in FIG. 3, and it is possible that not all operations in FIG. 3 are to be performed.

301. Receive a first image.

Specifically, the image sensor 13 of FIG. 1 can be used to acquire an image. Specifically, the lens 12 projects light reflected by a specific scene to the image sensor 13, and the image sensor 13 can transmit the digital image signal to the pre-processing module 14, and the pre-processing module 14 performs dead-point correction on the digital image signal, black. An operation such as level compensation is performed, and then the digital image after the pre-processing is input to the color shading estimation module 15. Here, the digital image signal may be referred to as image or image data.

302. Determine a tone value of each pixel in the first image.

Here, the tone value includes the rg tone component h ^rg and the bg tone component h ^bg . Specifically, the color shading estimation module 15 can obtain the rg hue component h ^rg and the bg hue component h ^bg of each pixel of the image according to the RGB values of the image received in 301.

303. Extract a first pixel point in the first image, where the first pixel point is a pixel point whose tone change is gentle.

和

其中

可以表示H ^rg的变化程度，

可以表示H ^bg的变化程度。 At this time, the logarithm operations of h ^rg and h ^bg of each pixel determined in 302 can be respectively performed to obtain H ^rg and H ^bg . The next step is to obtain a gradient for H ^rg and H ^bg respectively.

with

among them

Can indicate the degree of change in H ^rg ,

It can indicate the degree of change in H ^bg .

确定图像中rg色调的色调变化平缓的像素点，根据

确定图像中bg色调的色调变化平缓的像素点。这里，可以称色调变化平缓的像素点为第一像素点，将第一像素点所属的区域称为第一区域。也就是说，第一区域中的第一像素点的色调变化小于或等于阈值。 At this time, according to

Determining the pixel point of the rg tone in the image with a gentle change in color, according to

A pixel point in which the hg of the bg tone in the image changes gently is determined. Here, a pixel point whose tone change is gentle may be referred to as a first pixel point, and an area to which the first pixel point belongs may be referred to as a first area. That is, the change in hue of the first pixel in the first region is less than or equal to the threshold.

Specifically, the first region of the rg hue can be expressed as R ^rg , then:

Where x is any one of R ^rg , ie the first pixel, and th _rg is the hue change threshold of the rg hue.

The first region of the bg hue can be expressed as R ^bg , then:

Where x is any one of R ^bg , ie the first pixel, and th _bg is the hue change threshold of bg hue.

It can be understood that because the scene corresponding to the first image acquired in 301 can be a specific scene in reality, the image can have rich colors. In the embodiment of the present application, if the change in the hue of the image is relatively large, it can be considered that the color of the region where the hue changes is relatively large is the color of the scene itself. If the hue change of the image is relatively flat, it can be considered that the color of the area where the hue changes gently is the color shading outline.

304. Determine a transform coefficient of the first pixel by using a tone value of the first pixel and a feature profile of the first pixel, where the feature profile is used to represent a plurality of standard light sources obtained by statistics A feature of a color shaded outline of the image, the transform coefficient being used to indicate the degree of strength of the feature profile.

Specifically, the color shadow estimation module 15 may determine the feature contour of the first region according to a set of feature contours of each pixel point of the pre-stored image and coordinates of the pixel points of the first region. Here, a set of feature contours of each pixel point stored in advance may be a feature profile determined according to the method shown in FIG. 2 described above.

Then, a transform coefficient of the first region is determined according to a tone value of the first pixel in the first region and a feature contour of the first pixel. Step 304 can be performed by color shading estimation module 15.

则有： At this time, the transform coefficient of the first region can be defined.

Then there are:

Wherein x represents the first pixel point, F _k (x) is the k-dimensional feature profile of x, A is a transform coefficient _k F _k (x) is, B _k is F _k (x) the translation factor And a _k + b _k =1, k ranges from 1 to N and represents the dimension of the feature profile, N is a positive integer greater than 1, and H(x) represents the logarithm of the hue value at x.

则有： Specifically, for the rg hue, the transform coefficient of the first region can be defined.

Then there are:

则有： For bg hue, the transform coefficient of the first region can be defined

Then there are:

The derivation process of equation (3) will be described below.

In particular, a color shaded outline is a color shade of the image that is characterized using a color shading model. In the embodiment of the present application, the feature contour of the first pixel point is substituted into the color shadow model, and a function expression of the color shadow contour of the first pixel point can be obtained. That is, the color shadow contour of the first pixel in the first region may be represented as a function of the transform coefficient of the first pixel, wherein the transform coefficient of the first pixel is an independent variable, first The color shadow outline of a pixel is a dependent variable.

The tone value is the actual image signal of the extracted image. At this point, the hue value of the first pixel point can be made equal to the color shadow outline of the first pixel point, namely:

Where h(x) is the hue value of the first region.

For the two sides of equation (4), first take the logarithm and then find the gradient, then you can get:

的取值为第一像素点的变换系数的最优值。具体而言，上述公式(3)即表示

为第一像素点的实际的色调值的对数的梯度与第一像素点的颜色阴影轮廓的对数的梯度的差值最小时

的取值。 It can be understood that the above equations (4) and (5) are established when the color shading model is an ideal model. In practical applications, the gradient of the logarithm of the tonal value of the image and the logarithm of the color shade represented by the color shading model are often not exactly equal. At this time, the closer the gradient of the logarithm of the hue value of the image to the gradient of the logarithm of the color shading represented by the color shading model, the more the color shading represented by the color shading model reflects the actual color shading of the image. Therefore, it can be determined that when the gradient of the logarithm of the actual tone value of the first pixel point is closest to the gradient of the logarithm of the color shadow profile of the first pixel point, the transform coefficient

The value of the value is the optimum value of the transform coefficient of the first pixel. Specifically, the above formula (3) means

When the difference between the gradient of the logarithm of the actual tone value of the first pixel point and the gradient of the logarithm of the color shadow profile of the first pixel point is the smallest

The value.

的取值并不限定于上述公式(3)，例如还可以确定变换系数

为第一像素点的实际的色调值的对数的梯度与第一像素点的颜色阴影轮廓的对数的梯度的比值最接近1时

的取值，本申请实施例对此不限定。 It should be understood that, in the embodiment of the present application, the transform coefficient

The value is not limited to the above formula (3), for example, the transform coefficient can also be determined.

The ratio of the gradient of the logarithm of the actual tone value of the first pixel to the gradient of the logarithm of the color shadow profile of the first pixel is closest to 1

The value of the application is not limited in this embodiment.

305. Determine, according to a transform coefficient and a color shadow model of the first pixel, a color shadow contour of a second pixel in the first image, where the color shadow model is used to represent a color shadow of the image. The mapping between contours, feature contours, and transform coefficients.

Here, the pixel point in the first image is referred to as the second pixel point. In the embodiment of the present application, the transform coefficient of the first pixel in the first region may be used as the transform coefficient of the second pixel in the acquired first image. That is to say, in the embodiment of the present application, the transform coefficients of the respective pixel points in the image are the same. At this time, the transform coefficient and the feature contour of the second pixel point can be brought into the image color shadow model to synthesize the color shadow outline of the image.

At this time, the color shadow outline s(x') of the second pixel point x' is in the following form:

Feature profile wherein, f _k (x ') is of x', a _k is f _k (x ') of transform coefficients, b _k is f _k (x') of the translation factor and a _k + b _k = 1, N is a positive integer greater than 1 and represents the number of the characteristic contours f _k (x'), and k ranges from 1 to N.

可以确定第一图像的rg色调的颜色阴影轮廓为： Specifically, for the rg hue, based on the calculated model parameters of the first image

The color shadow outline of the rg hue of the first image can be determined as:

可以确定第一图像的bg色调的颜色阴影轮廓为： For the bg hue, based on the calculated model parameters of the first image

It can be determined that the color shadow outline of the bg tone of the first image is:

S306. Correct color shading of the second pixel point according to a color shadow profile of the second pixel point.

Specifically, the correction module 16 may maintain the correction value in the correction table according to the color shading profile of the image determined by the color shading estimation module 15. Specifically, the correction table may include an image pixel point and a correction value of the pixel point, and the correction value of the pixel point may specifically be a reciprocal of the color shadow profile of the pixel point.

Specifically, the correction value g ^rg (x') of the rg hue component of the second pixel point can be expressed as:

g ^rg (x')=1/s ^rg (x');

The correction value g ^bg (x') of the bg hue of the second pixel can be expressed as:

g ^bg (x')=1/s ^bg (x');

Specifically, the correction module 16 can correct the color shading of the first image according to the correction table.

Specifically, the R channel is corrected using a correction table of rg tone, namely:

r ^c (x')=r(x')·g ^rg (x') (9)

Where r(x') is the value of the R channel of the second pixel point x' before correction, and r ^c (x') is the value of the R channel of the second pixel point x' after correction.

The B channel is corrected using a bg tone correction table, ie:

b ^c (x')=b(x)·g ^bg (x') (10)

Where b(x') is the value of the B channel of the second pixel point x' before correction, and b ^c (x') is the value of the B channel of the second pixel point x' after correction.

After the correction module 16 corrects the color shading of the first region based on the correction table, the correction module 16 passes the image data to the subsequent processing module 17 for image processing.

The embodiment of the present application first extracts a tone value of the first pixel in the first region of the first image, and then determines the first pixel according to the tone value of the first pixel and the feature contour of the first pixel. Model parameters (ie, transform coefficients), further the model parameters and color shading models of the first pixel, determining a color shading contour of the first image, and finally performing color shading on the first image according to the color shading contour of the first image Correction. Since the embodiment of the present application determines the model parameters by directly using the tonal value of the image, and does not rely on other modules such as white balance to acquire the model parameters, the embodiment of the present application can accurately determine the model parameters. Moreover, in the embodiment of the present application, the feature contour can represent the color shadow caused by a specific factor or a statistical feature, and the color shadow model can accurately represent the complex color shadow contour, and thus the embodiment of the present application can accurately image the color shadow of the image. Make corrections.

FIG. 4 shows a schematic block diagram of a color shading correction apparatus 400 in accordance with an embodiment of the present application. The device 400 can be a cell phone, a digital camera, a tablet, a personal digital assistant, or other device that can capture images, or a desktop computer, video conferencing station, or other device that captures images using internal or external cameras. Specifically, the apparatus 400 includes a receiving unit 410, an extracting unit 420, a determining unit 430, and a correcting unit 440.

The receiving unit 410 is configured to receive the first image.

The extracting unit 420 is configured to extract a tone value of the first pixel in the first region in the first image, wherein a gradient of the tone value of the first pixel is less than or equal to a threshold.

a determining unit 430, configured to determine, by using a tone value of the first pixel point and a feature contour of the first pixel point, a transform coefficient of the first pixel point, where the feature contour is used to represent a statistically obtained A feature of a color shaded outline of a pixel point in an image of a plurality of standard light sources, the transform coefficient being used to indicate a degree of strength of the feature outline.

The determining unit 430 is further configured to determine, according to the transform coefficient and the color shadow model of the first pixel, a color shadow contour of the second pixel in the first image, where the color shadow model is used to represent The mapping between the color shadow outline, the feature outline, and the transform coefficients of the pixels in the image.

The correcting unit 440 is configured to correct a color shadow of the second pixel point according to a color shadow profile of the second pixel point.

The embodiment of the present application first extracts a tone value of the first pixel in the first region of the first image, and then determines the first pixel according to the tone value of the first pixel and the feature contour of the first pixel. Model parameters (ie, transform coefficients), further the model parameters and color shading models of the first pixel, determining a color shading contour of the first image, and finally performing color shading on the first image according to the color shading contour of the first image Correction. Since the embodiment of the present application determines the model parameters by directly using the tonal value of the image, and does not rely on other modules such as white balance to acquire the model parameters, the embodiment of the present application can accurately determine the model parameters. Moreover, in the embodiment of the present application, the feature contour can represent the color shadow caused by a specific factor or a statistical feature, and the color shadow model can accurately represent the complex color shadow contour, and thus the embodiment of the present application can accurately image the color shadow of the image. Make corrections.

Optionally, the extracting unit 420 is further configured to extract a color shadow profile of a pixel point of the flat field image of the plurality of standard light sources. The determining unit 430 is further configured to determine a feature vector of a color shadow contour of a pixel point of the flat field image of the plurality of standard light sources, and determine the feature contour according to the feature vector.

Optionally, the determining unit 430 is specifically configured to:

Determining the transform coefficients of the first pixel according to the following formula:

为x的变换系数且

f _k(x)为x的第k维的特征轮廓，a _k为f _k(x)的变换系数，b _k为f _k(x)的平移系数，且a _k+b _k＝1，k的取值范围为1至N且代表所述特征轮廓的维度，N为大于1的正整数，H(x)表示x处的色调值的对数。 Where x represents the first pixel point,

Is the transform coefficient of x and

F _k (x) is the k-dimensional feature profile of x, A _k is a transform coefficient F _k (x) is, B _k is F _k (x) the translation factor and _{a k + b k = 1,} k is The value ranges from 1 to N and represents the dimension of the feature profile, N is a positive integer greater than 1, and H(x) represents the logarithm of the tone value at x.

Alternatively, the color shading model can have the following form:

Wherein, s (x) denotes a color shading profile of x, f _k (x) for the first K x dimension of the feature profile, a _k is f _k (x) of transform coefficients, b _k is f _k (x) is translated The coefficient, k, ranges from 1 to N and represents the number of features, and N is a positive integer greater than one. In the embodiment of the present application, it is agreed that the colorless shadow f _k (x)=1, in order to maintain the normalized property of f _k (x), a _k +b _k =1.

In the embodiment of the present application, the transform coefficient a _{k is} used to describe the degree of strength of the feature contour f _k (x) corresponding to the transform coefficient a _k . Therefore, the color shading model in the embodiment of the present application can characterize a plurality of factors or color shading components when statistical features are superimposed by multiplying a plurality of feature contours.

Optionally, the color shadow profile of the second pixel determined by the determining unit 430 is:

Wherein, x 'representing the second pixel, s (x') represented by x 'shadow color profile, f _k (x') as x 'of the k-dimensional feature profile, a _k is F _k (x' The transform coefficient, b _k is the translation coefficient of f _k (x'), and a _k +b _k =1, k ranges from 1 to N and represents the number of features, N is greater than 1 Positive integer.

The correction unit is specifically configured to:

Determining a reciprocal of a color shadow profile of the second pixel point as a correction value of the second pixel point;

Correcting the color shading of the second pixel point according to the correction value of the second pixel point. It should be noted that, in the embodiment of the present invention, the receiving unit 410, the extracting unit 420, the determining unit 430, and the correcting unit 440 may be implemented by a processor. As shown in FIG. 5, the color shading correction device 600 can include a processor 610 and a memory 620. The memory 620 can be used to store feature contours and code executed by the processor 610, and the like.

In the implementation process, each step of the foregoing method may be completed by an integrated logic circuit of hardware in the processor 610 or an instruction in a form of software. The steps of the method disclosed in the embodiments of the present invention may be directly implemented as a hardware processor, or may be performed by a combination of hardware and software modules in the processor. The software module can be located in a conventional storage medium such as random access memory, flash memory, read only memory, programmable read only memory or electrically erasable programmable memory, registers, and the like. The storage medium is located in the memory 620, and the processor 610 reads the information in the memory 620 and completes the steps of the above method in combination with its hardware. To avoid repetition, it will not be described in detail here.

The color shading correction device 400 shown in FIG. 4 or the color shading correction device 600 shown in FIG. 5 can implement the respective processes corresponding to the method embodiments shown in FIG. 2 and FIG. 3, specifically, the color shading correction. For the device 400 or the color shading correction device 600, reference may be made to the descriptions in FIG. 2 and FIG. 3 above. To avoid repetition, details are not described herein again.

It should be understood that, in the various embodiments of the present application, the size of the sequence numbers of the foregoing processes does not mean the order of execution sequence, and the order of execution of each process should be determined by its function and internal logic, and should not be applied to the embodiment of the present application. The implementation process constitutes any limitation.

The embodiment of the present application further provides a computer readable medium for storing a computer program, the computer program comprising instructions for executing a corresponding method in the foregoing method embodiment.

The embodiment of the present application further provides a computer program product, comprising: computer program code, when the computer program code is run by a processor of a device for color shading correction, causing the color shading correction device to execute The corresponding method in any of the above method embodiments.

Those of ordinary skill in the art will appreciate that the elements and algorithm steps of the various examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the solution. A person skilled in the art can use different methods to implement the described functions for each particular application, but such implementation should not be considered to be beyond the scope of the present application.

A person skilled in the art can clearly understand that for the convenience and brevity of the description, the specific working process of the system, the device and the unit described above can refer to the corresponding process in the foregoing method embodiment, and details are not described herein again.

In the several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. For example, the device embodiments described above are merely illustrative. For example, the division of the unit is only a logical function division. In actual implementation, there may be another division manner, for example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.

The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

The functions may be stored in a computer readable storage medium if implemented in the form of a software functional unit and sold or used as a standalone product. Based on such understanding, the technical solution of the present application, which is essential or contributes to the prior art, or a part of the technical solution, may be embodied in the form of a software product, which is stored in a storage medium, including The instructions are used to cause a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like, which can store program codes. .

The foregoing is only a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto, and any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application. It should be covered by the scope of protection of this application. Therefore, the scope of protection of the present application should be determined by the scope of the claims.

Claims (12)

A method for color shading correction, comprising: Receiving a first image; Extracting a tone value of the first pixel in the first image, wherein a gradient of a tone value of the first pixel is less than or equal to a threshold; Determining a transform coefficient of the first pixel point by using a tone value of the first pixel point and a feature profile of the first pixel point, wherein the feature profile is used to represent an image of a plurality of statistically derived standard light sources a feature of a color shaded outline of a pixel in which the transform coefficient is used to indicate a degree of strength of the feature outline; Determining a color shadow profile of a second pixel point in the first image according to a transform coefficient and a color shadow model of the first pixel, wherein the color shadow model is used to represent a color shadow of a pixel point in the image a mapping relationship between contours, feature contours, and transform coefficients; Correcting the color shading of the second pixel point according to the color shading profile of the second pixel point. The method according to claim 1, wherein the determining a transform coefficient of the first pixel by using a tone value of the first pixel and a feature profile of the first pixel includes: Determining the transform coefficients of the first pixel according to the following formula:

Where x represents the first pixel point,

Is the transform coefficient of x and

F _k (x) is the k-dimensional feature profile of x, A _k is a transform coefficient F _k (x) is, B _k is F _k (x) the translation factor and _{a k + b k = 1,} k is The value ranges from 1 to N and represents the dimension of the feature profile, N is a positive integer greater than 1, and H(x) represents the logarithm of the tone value at x. The method according to claim 1 or 2, wherein the color shadow outline of the second pixel point is:

Wherein, x 'representing the second pixel, s (x') represented by x 'shadow color profile, f _k (x') as x 'of the k-dimensional feature profile, a _k is F _k (x' The transform coefficient, b _k is the translation coefficient of f _k (x'), and a _k +b _k =1, k ranges from 1 to N and represents the number of features, N is greater than 1 Positive integer. The method according to any one of claims 1 to 3, wherein the correcting the color shading of the second pixel point according to the color shading contour of the second pixel point comprises: Determining a reciprocal of a color shadow profile of the second pixel point as a correction value of the second pixel point; Correcting the color shading of the second pixel point according to the correction value of the second pixel point. The method according to any one of claims 1 to 4, wherein the determining the transformation of the first pixel point by using a tone value of the first pixel point and a feature contour of the first pixel point Before the coefficient, it also includes: Extracting a color shadow outline of a pixel point of the flat field image of the plurality of standard light sources; Determining a feature vector of a color shadow profile of a pixel of the flat field image of the plurality of standard light sources, and determining the feature profile based on the feature vector. A device for color shading correction, comprising: a receiving unit, configured to receive the first image; An extracting unit, configured to extract a tone value of a first pixel in the first image, wherein a gradient of a tone value of the first pixel is less than or equal to a threshold; a determining unit, configured to determine, by using a tone value of the first pixel point and a feature contour of the first pixel point, a transform coefficient of the first pixel point, wherein the feature contour is used to represent a statistically obtained a feature of a color shaded outline of a pixel in an image of a standard light source, the transform coefficient being used to indicate a degree of strength of the feature outline; The determining unit is further configured to determine a color shadow profile of the second pixel point in the first image according to the transform coefficient and the color shadow model of the first pixel, wherein the color shadow model is used to represent the image a color shadow contour, a feature contour, and a mapping relationship between transform coefficients of the pixel in the pixel; And a correcting unit, configured to correct a color shadow of the second pixel point according to a color shadow profile of the second pixel point. The device according to claim 6, wherein the determining unit is specifically configured to: Determining the transform coefficients of the first pixel according to the following formula:

Where x represents the first pixel point,

Is the transform coefficient of x and

F _k (x) is the k-dimensional feature profile of x, A _k is a transform coefficient F _k (x) is, B _k is F _k (x) the translation factor and _{a k + b k = 1,} k is The value ranges from 1 to N and represents the dimension of the feature profile, N is a positive integer greater than 1, and H(x) represents the logarithm of the tone value at x. The apparatus according to claim 6 or 7, wherein the determining unit determines a color shading profile of the second pixel point as:

Wherein, x 'representing the second pixel, s (x') represented by x 'shadow color profile, f _k (x') as x 'of the k-dimensional feature profile, a _k is F _k (x' The transform coefficient, b _k is the translation coefficient of f _k (x'), and a _k +b _k =1, k ranges from 1 to N and represents the number of features, N is greater than 1 Positive integer. The device according to any one of claims 6-8, wherein the correcting unit is specifically configured to: Determining a reciprocal of a color shadow profile of the second pixel point as a correction value of the second pixel point; Correcting the color shading of the second pixel point according to the correction value of the second pixel point. The device according to any one of claims 6-9, further comprising: The extracting unit is further configured to extract a color shadow profile of a pixel point of the flat field image of the plurality of standard light sources; The determining unit is further configured to determine a feature vector of a color shadow profile of a pixel point of the flat field image of the plurality of standard light sources, and determine the feature contour according to the feature vector. A color shading correction apparatus, comprising: a memory and a processor, wherein the memory is for storing an instruction, the processor is configured to execute the memory stored instruction, so that the processor executes as A method as claimed in any one of claims 1 to 5. A computer readable medium for storing a computer program, the computer program comprising instructions for performing the method of any of claims 1-5.

Images