It should be noted that the essence of performing the radial chromatic aberration processing on the first image is to recalculate the color values of individual pixels on the first image, and after the individual pixels are assigned with the new color values, an image with the radial chromatic aberration effect is obtained. Specifically, the procedure of the radial chromatic aberration processing may include S304-S305.

S304. obtaining, for each pixel on the first image, a sum of color values of a plurality of sampling points corresponding to the pixel in each color channel based on the coordinates of the trigger pixel, coordinates of the pixel, a number of the sampling points, a step coefficient, an intensity coefficient corresponding to the color channel, a weight coefficient, a texture of the first image, and the chromatic aberration intensity coefficient.

In one of the possible implementations, the sum of color values of the plurality of sampling points corresponding to the pixel in each color channel may be obtained by the following steps. Taking any of the pixels on the first image as an example, for the purpose of illustration, the pixel is referred to as the current pixel in the present disclosure, as shown in FIG. 4, it specifically includes:

S304-A. determining a direction from the trigger pixel to the current pixel based on the coordinates of the trigger pixel and the coordinates of the current pixel.

S304-B. determining a sampling step based on the direction from the trigger pixel to the current pixel, the step coefficient, and the number of the sampling points.

Specifically, the sampling step may be determined by the following equation:

step
=

dir
*
radius
⁢

Strength
*
u_Sample

where step represents the sampling step, dir represents the direction from the trigger pixel to the current pixel, radiusStrength represents the step coefficient, u_Sample represents the number of the sampling points, where the step coefficient and the number of the sampling points may be predetermined values.

S304-C. determining, for each color channel of RGB channels, an offset corresponding to the color channel based on the direction from the trigger pixel to the current pixel, the chromatic aberration intensity coefficient and the intensity coefficient corresponding to the color channel.

Specifically, the offset corresponding to the red channel may be determined by the following equation:

red
⁢

Offset

=

dir
*
u_Strength
*
u_Red
⁢

Strength

where redOffset represents the offset corresponding to the red channel, dir represents the direction from the trigger pixel to the current pixel, u_Strength represents the chromatic aberration intensity coefficient, which may be obtained by S302, u_RedStrength represents the intensity coefficient corresponding to the red channel, which may be a predetermined value.

The offset corresponding to the green channel may be determined by the following equation:

green
⁢

Offset

=

dir
*
u_Strength
*
u_Green
⁢

Strength

where greenOffset represents the offset corresponding to the green channel, dir represents the direction from the trigger pixel to the current pixel, u_Strength represents the chromatic aberration intensity coefficient, which may be obtained by S302, u_GreenStrength represents the intensity coefficient corresponding to the green channel, which may be a predetermined value.

The offset corresponding to the blue channel may be determined by the following equation:

blue
⁢

Offset

=

dir
*
u_Strength
*
u_Blue
⁢

Strength

where blueOffset represents the offset corresponding to the blue channel, dir represents the direction from the trigger pixel to the current pixel, u_Strength represents the chromatic aberration intensity coefficient, which may be obtained by S302, u_BlueStrength represents the intensity coefficient corresponding to the blue channel, which may be a predetermined value.

S304-D. determining, for each color channel of the RGB channels, the sum of the color values of the plurality of sampling points corresponding to the current pixel in the color channel based on the texture of the first image, the coordinates of the current pixel, the offset corresponding to the color channel, the sampling step, the number of the sampling points and the weight coefficient.

Specifically, for any of the color channels, a loop statement may be used to determine the sum of the color values of a plurality of sampling points corresponding to the current pixel in the color channel. Taking the red channel as an example, the formula in the loop statement is as below:

R
⁢

+
=

⁢

texture
⁢

⁢
2
⁢

D
⁡
(

Input
⁢

Texture

,

uv
+

red
⁢

Offset

)

.
r

*
weight

Each time a loop is completed, the offset corresponding to the red channel is subtracted by one step, and the number of loops is equal to the number of sampling points. R represents the sum of the color values of the plurality of sampling points in the red channel, InputTexture represents the texture of the first image, uv represents the coordinates of the current pixel point, redOffset represents the offset corresponding to the red channel, weight represents the weight coefficient, which may be a predetermined value.

S305. determining a color value of the current pixel in each color channel based on the sum of the color values of the plurality of sampling points corresponding to the current pixel in the color channel.

For each channel, the color value of the current pixel in the color channel may be obtained from the following formula, which takes the red channel as an example: R/=u_Sample, where R represents the sum of the color values of a plurality of sampling points corresponding to the current pixel in the red channel, and u_Sample represents the number of sampling points.

An example is given below:

Referring to FIG. 5, assuming that the pixel O is the trigger pixel, the pixel M is the current pixel, and a plurality of sampling points corresponding to the current pixel are M1, M2 and M3, the sum of the color values of M1, M2 and M3 in channel R is obtained by the above method as R1+R2+R3, and the sum of the color values of M1, M2 and M3 in channel G is obtained by the above method as G1+G2+G3, the sum of color values of M1, M2 and M3 in channel B is B1+B2+B3, then the RGB values of M may be determined as: (R1+R2+R3)/3, (G1+G2+G3)/3, (B1+B2+B3)/3.

By performing the processing as S304-S305 above on all pixels on the first image, the RGB values of all pixels can be obtained. By assigning the calculated RGB values to the corresponding pixel points, the image processed by the radial chromatic aberration processing can be obtained.

The method for image processing provided in the embodiment provides a method of radial chromatic aberration processing, and the image processed by applying this method has the effect of radial chromatic aberration in a real magnifying glass, and the effect is highly realistic.

The Third Embodiment

FIG. 6 shows a schematic flowchart of the third embodiment of the method for image processing provided by the present disclosure. As described above, the effect processing in the present disclosure may include distortion processing, and this embodiment illustrates a procedure of the distortion processing. As shown in FIG. 6, the method for image processing provided in this embodiment includes:

S601. in response to a first trigger operation from a user, obtaining screen coordinates of the first trigger operation.

S602. obtaining, for each first image obtained after the first trigger operation, coordinates of a trigger pixel on the first image based on the screen coordinates of the first trigger operation.

The implementation process of S601 and S602 refer to the above description, and will not be repeated here in the present application.

S603. obtaining a distortion coefficient corresponding to the first image.

As described above, the first mapping may be established in advance, which is used for indicating the correspondence between the frame number and the distortion coefficient. In processing the first image, the distortion coefficient corresponding to the first image is first determined based on the frame number of the first image and the first mapping, and then, based on the coordinates of the trigger pixel, the effect processing is performed on the first image by using the distortion coefficient corresponding to the first image.

The procedure of the distortion processing will be introduced below. It should be noted that as described above, the essence of the radial chromatic aberration processing is to recalculate the color values of individual pixels on the first image. Analogously, the essence of the distortion processing is to recalculate the color values of individual pixels on the first image. After the individual pixels are assigned with the new color values, an image with the distortion effect is obtained. Specifically, S604-S607 are included.

S604. obtaining a distortion function based on the distortion coefficient.

In one of the possible implementations, assuming that f(x)=(k−1)x²+x, all functions with the value of k satisfying 0.5≤k≤1.0 can be used as the distortion function, and k can be used as the distortion coefficient, Assuming that the value of k obtained in S603 is 0.75, then the distortion function is f(x)=(0.75−1)x²+x=−0.25x²+x.

S605. determining, for each pixel on the first image, a pixel before distortion on the first image based on the coordinates of the trigger pixel, the coordinates of the pixel, a distance from the trigger pixel to the pixel and the distortion function, the pixel before distortion corresponding to the pixel.

Specifically, taking any of the pixels on the first image as an example, for the purpose of illustration, the pixel is referred to as the current pixel in the present disclosure. The pixel before distortion corresponding to the current pixel may be determined through the formula as below:

uv
=

(

texture
⁢

Coordinate

-
center

)

*

(

f
⁡
(
dis
)

dis

)

+
center

where uv represents the coordinates of the pixel before distortion corresponding to the current pixel, textureCoordinate represents the coordinates of the current pixel, center represents the coordinates of the trigger pixel, dis represents the distance from the trigger pixel to the current pixel, and f represents the distortion function.

S606. determining a color value of the pixel before distortion as the color value of the current pixel.

After the coordinates of the pixel before distortion corresponding to the current pixel are determined through the above S605, the pixel before distortion can be found on the first image, and the color value of the pixel before distortion is determined as the color value of the current pixel.

An example is given below:

FIG. 7 shows the image processed by the radial chromatic aberration processing. Assuming that the pixel O is the trigger pixel, the pixel M is the current pixel, the pixel before distortion corresponding to the pixel M that is obtained by S605 is M1, and the RGB values of M1 are R1, G1, and B1, then the RGB values of pixel M may be determined as R1, G1, and B1.

By performing the same processing as S605-S606 on all pixels on the first image, the color values of all pixels can be obtained. By assigning the calculated color values to the individual pixels, the image processed by the distortion processing can be obtained.

The method for image processing provided in the embodiment provides a method of distortion processing, and the image processed by applying this method has the distortion effect in a real magnifying glass, and the effect is highly realistic.

The Fourth Embodiment

FIG. 8 shows a schematic flowchart of the fourth embodiment of the method for image processing provided by the present disclosure. As described above, the effect processing in the present disclosure may include scaling processing, and this embodiment illustrates a procedure of the scaling processing. As shown in FIG. 8, the method for image processing provided in this embodiment includes:

S801. in response to a first trigger operation from a user, obtaining screen coordinates of the first trigger operation.

S802. obtaining, for each first image obtained after the first trigger operation, coordinates of a trigger pixel on the first image based on the screen coordinates of the first trigger operation.

The implementation process of S801 and S802 refer to the above description, and will not be repeated here in the present application.

S803. obtaining the scaling coefficient corresponding to the first image.

As described above, the first mapping may be established in advance, which is used for indicating the correspondence between the frame number and the scaling coefficient. In processing the first image, the scaling coefficient corresponding to the first image is first determined based on the frame number of the first image and the first mapping, and then, based on the coordinates of the trigger pixel, the effect processing is performed on the first image by using the scaling coefficient corresponding to the first image.

The procedure of the scaling processing will be introduced below, which specifically includes S804-S806.

S804. determining scaled vertex coordinates based on the coordinates of the trigger pixel, current vertex coordinates of a quadrilateral model 10 and the scaling coefficient, the quadrilateral model 10 being used for changing a display size of the image.

Specifically, the scaled vertex coordinates may be calculated through the formula as below:

pos
⁢
1

=

(

pos
-
center

)

*
scale

+
center

where pos represents the current vertex coordinates of the quadrilateral model 10, center represents the coordinates of the trigger pixel, scale represents the scaling coefficient, pos1 represents the scaled vertex coordinates.

S805. updating the vertex coordinates of the quadrilateral model 10 to the scaled vertex coordinates.

An example is given below:

As shown in FIG. 9, assuming that the pixel O is the trigger pixel, the current vertex coordinates of the quadrilateral model 10 are A, B, C and D, the scaling coefficient obtained by S803 is 1, it can be determined that the quadrilateral model 10 needs to be doubled with the point O as the center, and the scaled vertices are represented by A′, B′, C′ and D′.

S806. mapping the first image to the quadrilateral model 10 to obtain the image with the magnifying glass effect.

The method for image processing provided in the embodiment provides a method of scaling processing, and the image processed by applying this method has the scaling effect in a real magnifying glass, and the effect is highly realistic.

The Fifth Embodiment

FIG. 10 is a schematic flowchart of the fifth embodiment of the method for image processing provided by the present disclosure. As described above, the effect processing in the present disclosure may include radial blur processing, and this embodiment illustrates a procedure of the radial blur processing. As shown in FIG. 10, the method for image processing provided in this embodiment includes:

S1001. in response to a first trigger operation from a user, obtaining screen coordinates of the first trigger operation.

S1002. obtaining, for each first image obtained after the first trigger operation, coordinates of a trigger pixel on the first image based on the screen coordinates of the first trigger operation.

The implementation process of S1001 and S1002 refer to the above description, and will not be repeated here in the present application.

S1003. obtaining the blur coefficient corresponding to the first image.

As described above, the first mapping may be established in advance, which is used for indicating the correspondence between the frame number and the blur coefficient. In processing the first image, the blur coefficient corresponding to the first image is first determined based on the frame number of the first image and the first mapping, and then, based on the coordinates of the trigger pixel, the effect processing is performed on the first image by using the blur coefficient corresponding to the first image.

Analogous to the radial chromatic aberration processing and the distortion processing, the essence of the radial blur processing is to recalculate the color values of individual pixels on the first image, and after the individual pixels are assigned with the new color values, an image with the radial blur effect is obtained. Specifically, S1004-S1005 are included.

S1004. obtaining, for each pixel on the first image, a sum of color values of a plurality of sampling points corresponding to the pixel based on the coordinates of the trigger pixel, coordinates of the pixel, a number of the sampling points, a texture of the first image, and the blur coefficient.

Taking any of the pixels on the first image as an example, for the purpose of illustration, the pixel is referred to as the current pixel in the present disclosure. The sum of color values of the plurality of sampling points corresponding to the current pixel may be determined through the following method:

First, the direction from the trigger pixel to the current pixel is determined based on the coordinates of the trigger pixel and the coordinates of the current pixel. The specific procedure refers to the above embodiments, and the details will not be repeated herein. The sum of the color values of the plurality of sampling points corresponding to the current pixel is determined based on the coordinates of the current pixel, a number of the sampling points, the blur coefficient, the texture of the first image, and the direction from the trigger pixel to the current pixel.

Specifically, a loop statement may be used to determine the sum of the color values of a plurality of sampling points corresponding to the current pixel in the color channel. The formula in the loop statement is as below:

vec
⁢
2
⁢
uv

=

uv
+

blur
⁢

factor
*
dir
*
i

;

out
⁢

Color

+=

texture
⁢

2
⁢

D
⁡
(

Input
⁢

Texture

,
uv

)

where i represents a loop variable, the number of loops is equal to the number of the sampling points, uv represents the coordinates of the current pixel, blurfactor represents the blur coefficient, dir represents the direction from the trigger pixel to the current pixel, InputTexture represents the texture of the first image, outColor represents the sum of the color values of the plurality of sampling points, vec 2 represents the coordinates of the current pixel, and uv represents a two-dimensional vector.

S1005. obtaining a color value of the current pixel based on the sum of the color values of the plurality of sampling points corresponding to the current pixel.

Specifically, the color value of the current pixel may be obtained by the following formula: outColor/=u_Sample, where outColor represents the sum of the color values of the plurality of sampling points corresponding to the current pixel, u_Sample represents the number of the sampling points.

By performing the processing as S1004-S1005 on all pixels on the first image, the color values of all pixels can be obtained. By assigning the calculated color values to the corresponding pixel points, the image processed by the radial blur processing can be obtained.

The method for image processing provided in the embodiment provides a method of radial blur processing, and the image processed by applying this method has the radial blur effect in a real magnifying glass, and the effect is highly realistic.

In one of the possible implementations, the radial chromatic aberration processing, the distortion processing, the scaling processing, and the radial blur processing may be performed in sequence on the first image based on the coordinates of the trigger pixel and the processing parameter corresponding to the first image. In such an implementation, the result after the radial chromatic aberration processing is the object of the distortion processing, that is, the first image in the procedure of the distortion processing as shown in FIG. 6 is the image after the radial chromatic aberration processing. The result of the distortion processing is the object of the scaling processing, that is, the first image in the procedure of the scaling processing as shown in FIG. 8 is the image processed by the distortion processing. The result of the scaling processing is the object of the radial blur processing, that is, the first image in the procedure of the radial blur as shown in FIG. 10 is the image processed by the scaling processing. The image obtained by such a processing sequence is closer to the effect of the real magnifying glass.

By way of example, in the first mapping, if the frame number may be positively correlated with the three following parameters, that is, the chromatic aberration intensity coefficient, the scaling coefficient and the blur coefficient respectively, and be negatively correlated with the distortion coefficient, the magnification degree of the image obtained by the above procedure of image processing becomes larger frame by frame. In one of the possible implementations, after the user triggers the first trigger operation on the shooting interface shown in FIG. 2, a second trigger operation may also be triggered; in response to the second trigger operation, screen coordinates of the second trigger operation are obtained; for each second image obtained after the second trigger operation, coordinates of a trigger pixel on the second image are obtained based on the screen coordinates of the second trigger operation; a processing parameter corresponding to the second image is obtained based on a frame number of the second image and second mapping; and effect processing is performed on the second image based on the coordinates of the trigger pixel on the second image and the processing parameter corresponding to the second image. In the second mapping, the frame number may be negatively correlated with the chromatic aberration intensity coefficient, the scaling coefficient and the blur coefficient respectively, and be positively correlated with the distortion coefficient. As such, after the terminal device receives the second trigger operation, the magnification degree of the image obtained by the above procedure of image processing becomes smaller frame by frame, so that the change of amplification degree of the image after the user triggers the first trigger operation and the change of amplification degree of the image after the user triggers the second trigger operation are two opposite processes, that is, the magnification degree becomes larger frame by frame after the user triggers the first trigger operation, while the magnification degree becomes smaller frame by frame after the user triggers the second trigger operation, thereby enhancing the user's enjoyment in making videos.

It should be understood that in the first mapping, the relationship between the frame number and the chromatic aberration intensity coefficient, the scaling coefficient and the blur coefficient may be configured to be negative correlation, and the relationship between the frame number and the distortion coefficient may be configured to be positive correlation. In the second mapping, the relationship between the frame number and the chromatic aberration intensity coefficient, the scaling coefficient and the blur coefficient may be configured to be positive correlation, and the relationship between the frame number and the distortion coefficient may be configured to be negative correlation. The result of such a configuration is that the magnification degree becomes smaller frame by frame after the user triggers the first trigger operation, and the magnification degree becomes larger frame by frame after the user triggers the second trigger operation.

FIG. 11 shows a schematic diagram of a structure of the apparatus for image processing provided by the present disclosure. As shown in FIG. 11, the apparatus for image processing provided by the present disclosure comprises:

- an obtaining module 1101 configured for in response to a first trigger operation from a user, obtaining screen coordinates of the first trigger operation;
- an effect processing module 1102 configured for obtaining, for each first image obtained after the first trigger operation, coordinates of a trigger pixel on the first image based on the screen coordinates of the first trigger operation; performing, based on the coordinates of the trigger pixel, effect processing on the first image to obtain an image with a magnifying glass effect.

Optionally, the effect processing module 1102 is specifically configured for:

- obtaining a processing parameter corresponding to the first image based on a frame number of the first image and first mapping, the processing parameter comprising at least one of a chromatic aberration intensity coefficient, a distortion coefficient, a scaling coefficient, and a blur coefficient, the first mapping is used for indicating a correspondence between the frame number and the processing parameter; and
- performing, based on the coordinates of the trigger pixel and the processing parameter corresponding to the first image, the effect processing on the first image to obtain the image with the magnifying glass effect,