Method and apparatus for rendering hair, computer storage medium, electronic device

Patent metadata

Field	Value
Publication number	US-12437450-B2
Application number	US-202318191018-A
Country	US
Kind code	B2
Filing date	Mar 28, 2023
Priority date	Apr 1, 2022
Publication date	Oct 7, 2025
Grant date	Oct 7, 2025

Abstract

A method for rendering hair is provided and includes: obtaining a first rendered result by performing transparent channel test rendering on a specified pixel point in a hair model, wherein the specified pixel point is a pixel point with a pixel depth value less than a preset depth threshold and a transparency channel value greater than a preset transparency threshold; obtaining a second rendered result by performing transparent channel hybrid rendering on a backlight pixel point of the hair model based on the first rendered result; obtaining a third rendered result by performing transparent channel hybrid rendering on a light-oriented pixel point of the hair model based on the second rendered result; obtaining a final rendered result of the hair model by hybridizing the first rendered result, the second rendered result and the third rendered result.

First claim

What is claimed is:

1. A method for rendering hair, performed by a server, comprising:
establishing a hair model in 3D software, wherein the hair model comprises a light-oriented pixel point and a backlight pixel point, the light-oriented pixel point is a pixel point directly facing a preset light source in a virtual scene, the backlight pixel point is a pixel point facing away from the preset light source in the virtual scene, and each pixel point in the hair model contains a color value and a transparent channel value;
obtaining a pixel depth value corresponding to each pixel point by rasterizing the hair model through a vertex shader;
obtaining a first rendered result by performing transparent channel test rendering on a first specified pixel point in the hair model through a fragment shader, wherein the transparent channel test rendering comprises performing rendering on the first specified pixel point with the pixel depth value less than a preset depth threshold and the transparency channel value greater than a preset transparency threshold;
obtaining a second rendered result by performing transparent channel hybrid rendering on the backlight pixel point of the hair model based on the first rendered result through the fragment shader;
obtaining a third rendered result by performing transparent channel hybrid rendering on the light-oriented pixel point of the hair model based on the second rendered result through the fragment shader,
wherein the transparent channel hybrid rendering comprises performing rendering on a second specified pixel point in the hair model, the second specified pixel point comprising the backlight pixel point or the light-oriented pixel point, and wherein the color value of the second specified pixel point is determined based on a following formula,
current color value of the second specified pixel point * transparent channel value corresponding to the second specified pixel point+color value of the second specified pixel point determined by previous rendering result * (1-transparent channel value corresponding to the second specified pixel point); and
obtaining a final rendered result of the hair model by hybridizing the first rendered result, the second rendered result and the third rendered result, wherein hybridizing the first rendered result, the second rendered result and the third rendered result comprises performing a weighted operation on the first rendered result, the second rendered result and the third rendered result, based on a preset weight.

2. The method according to claim 1 , wherein obtaining the first rendered result by performing the transparent channel test rendering on the first specified pixel point in the hair model through the fragment shader, comprises:
determining a half-angle vector according to a direction vector of the preset light source and a direction vector of a view angle of a camera, wherein the determining the half-angle vector according to the direction vector of the preset light source and the direction vector of the view angle of the camera, comprises:
acquiring a sum vector of the direction vector of the preset light source and the direction vector of the view angle of the camera; and
obtaining the half-angle vector by normalizing the sum vector;

determining a sine index according to a tangent vector of the specified pixel point and the half-angle vector, wherein the sine index is calculated by a following formula:
sin( T,H )=√{square root over (1- dot ( T,H ) 2 )}

where sin (T, H) denotes the sine index, T denotes the tangent vector, and H denotes the half-angle vector, dot (T, H) denotes a dot product of T and H;
determining a highlight color of the specified pixel point according to the sine index and a highlight intensity corresponding to the specified pixel point, wherein the determining the highlight color of the specified pixel point according to the sine index and the highlight intensity corresponding to the specified pixel point comprises:
performing an exponentiation operation by using the sine index as a base number and using the highlight intensity corresponding to the specified pixel point as an exponent; and
determining a result of the exponentiation operation as the highlight color of the specified pixel point; and

obtaining the first rendered result by performing transparent channel test rendering on the specified pixel point according to the highlight color through the fragment shader.

3. The method according to claim 1 , wherein the obtaining the final rendered result by hybridizing the first rendered result, the second rendered result and the third rendered result comprises:
obtaining a first hybrid result by determining a weighted sum of the first rendered result and the second rendered result based on the preset weight; and
obtaining the final rendered result by determining a weighted sum of the first hybrid result and the third rendered result based on the preset weight.

4. The method according to claim 1 , wherein obtaining the final rendered result by hybridizing the first rendered result, the second rendered result and the third rendered result comprises:
obtaining a second hybrid result by determining a weighted sum of the second rendered result and the third rendered result based on the preset weight; and
obtaining the final rendered result by determining a weighted sum of the second hybrid result and the first rendered result based on the preset weight.

5. The method according to claim 1 , wherein obtaining the second rendered result by performing the transparent channel hybrid rendering on the backlight pixel point of the hair model based on the first rendered result through the fragment shader, comprises:
determining a half-angle vector according to a direction vector of the preset light source and a direction vector of a view angle of a camera, wherein the determining the half-angle vector according to the direction vector of the preset light source and the direction vector of the view angle of the camera, comprises:
acquiring a sum vector of the direction vector of the preset light source and the direction vector of the view angle of the camera; and
obtaining the half-angle vector by normalizing the sum vector;

determining a sine index according to a tangent vector of the backlight pixel point and the half-angle vector, wherein the sine index is calculated by a following formula:
sin( T,H )=√{square root over (1- dot ( T,H ) 2 )}

where sin (T, H) denotes the sine index, T denotes the tangent vector, and H denotes the half-angle vector, dot (T, H) denotes a dot product of T and H;
determining a highlight color of the backlight pixel point according to the sine index and a highlight intensity corresponding to the backlight pixel point, wherein the determining the highlight color of the backlight pixel point according to the sine index and the highlight intensity corresponding to the backlight pixel point comprises:
performing an exponentiation operation by using the sine index as a base number and using the highlight intensity corresponding to the backlight pixel point as an exponent; and
determining a result of the exponentiation operation as the highlight color of the backlight pixel point; and

obtaining the second rendered result by performing transparent channel hybrid rendering on the backlight pixel point based on a following formula through the fragment shader:
the highlight color of the backlight pixel point * transparent channel value corresponding to the backlight pixel point+color value of the backlight pixel point determined by the first rendered result * (1-transparent channel value corresponding to the backlight pixel point).

6. The method according to claim 1 , wherein obtaining the third rendered result by performing the transparent chan

Assignees

• Beijing Wodong Tianjun Information Technology Co Ltd
• Beijing Jingdong Century Trading Co Ltd

Inventors

• Yang Chao
• Liu Xiangjun
• Du Feng
• Wu Chaoyang

Classifications

• G06T11/10 Primary

  Texturing; Colouring; Generation of textures or colours (retouching, inpainting or scratch removal G06T5/77) · CPC title

• G06T15/005

  General purpose rendering architectures · CPC title

• G06T2210/62

  Semi-transparency · CPC title

• G06T15/503

  Blending, e.g. for anti-aliasing · CPC title

• G06T15/506

  Illumination models · CPC title