Method for processing images and electronic device

Patent metadata

Field	Value
Publication number	US-11488293-B1
Application number	US-202217568156-A
Country	US
Kind code	B1
Filing date	Jan 4, 2022
Priority date	Apr 30, 2021
Publication date	Nov 1, 2022
Grant date	Nov 1, 2022

Abstract

A method for processing images is provided. The method includes: acquiring a first image by smoothing a skin region of a target object in an original image; determining a skin texture material matching with a face area of the target object; acquiring a facial texture image of the target object by rendering the skin texture material; and acquiring a second image by fusing the facial texture image with the first image.

First claim

What is claimed is:

1. A method for processing images, comprising:
acquiring a first image by smoothing a skin region of a target object in an original image;
determining a skin texture material matching with a face area of the target object, wherein the skin texture material is selected from pre-stored candidate skin texture materials, wherein the pre-stored candidate skin materials are texture materials for standard skin of a face of a character drawn by a technician;
acquiring a facial texture image of the target object by rendering the skin texture material, wherein both facial key point information and facial posture information of the facial texture image are matched with the target object; and
acquiring a second image by fusing the facial texture image with the first image.

2. The method according to claim 1 , wherein said determining the skin texture material matching with the face area of the target object comprises:
determining, based on a face area range within which the face area is, a resolution range matching with the face area range, wherein a median value of the face area range is positively correlated with a median value of the resolution range; and
acquiring the skin texture material whose resolution is within the resolution range.

3. The method according to claim 1 , wherein said acquiring the facial texture image of the target object by rendering the skin texture material comprises:
acquiring facial key point information and facial posture information of the target object, wherein the facial posture information of the target object is configured to indicate a face rotation condition of the target object;
acquiring a target texture material by rendering the skin texture material based on the facial key point information and the facial posture information of the target object; and
acquiring the facial texture image by fusing the target texture material with the original image.

4. The method according to claim 3 , wherein said acquiring the target texture material by rendering the skin texture material based on the facial key point information and the facial posture information of the target object comprises:
acquiring standard key point information of the skin texture material;
determining, based on the facial key point information of the target object and the standard key point information, a correspondence relationship between standard key points of the skin texture material and facial key points of the target object; and
acquiring the target texture material by performing texture mapping on the skin texture material based on the correspondence relationship and the facial posture information of the target object.

5. The method according to claim 1 , wherein said acquiring the first image by smoothing the skin region of the target object in the original image comprises:
acquiring facial key point information and facial posture information of the target object, wherein the facial posture information of the target object is configured to indicate a face rotation condition of the target object;
acquiring a posture weight image of the target object based on the facial posture information of the target object, wherein a pixel value of each pixel in the posture weight image is configured to indicate a posture weight parameter of a corresponding pixel in the original image, the posture weight parameter being configured to indicate an importance of the corresponding pixel relative to the facial posture information; and
acquiring the first image by smoothing the skin region based on the posture weight image.

6. The method according to claim 5 , wherein the facial posture information of the target object comprises an Euler angle of a facial posture of the target object, and said acquiring the posture weight image of the target object based on the facial posture information of the target object comprises:
acquiring a face orientation mask map of the target object based on a value symbol of the Euler angle of the facial posture, wherein the face orientation mask map is configured to indicate whether an orientation of a face of the target object is forward or backward;
acquiring a distance from each pixel in a face region of the target object to a face midline; and
acquiring the posture weight image based on the face orientation mask map and the distance.

7. The method according to claim 6 , wherein said acquiring the posture weight image based on the face orientation mask map and the distance comprises:
acquiring a first value by multiplying the distance for each pixel in the face region of the target object by a pixel value of a corresponding pixel in the face orientation mask map;
acquiring a second value by multiplying the first value by a first coefficient, wherein the first coefficient is an adjustment factor for a distance from a pixel to a face midline, and is greater than or equal to 0 and less than or equal to 1; and
acquiring a pixel value of a corresponding pixel in the posture weight image by calculating a difference between a second coefficient and the second value, wherein the second coefficient is equal to 1.

8. The method according to claim 5 , wherein said acquiring the first image by smoothing the skin region based on the posture weight image comprises:
acquiring a skin region image of the original image, wherein the skin region image is configured to indicate the skin region of the target object in the original image;
acquiring a smoothed skin tone image and a smoothed posture weight image by smoothing the skin region image and the posture weight image based on a two-dimensional Gaussian blur function; and
acquiring the first image by fusing the original image, the smoothed skin tone image, and the smoothed posture weight image.

9. The method according to claim 1 , further comprising:
acquiring a hair region image and an occluded face region image of the original image, and an average brightness parameter of a face region of the target object in the original image, wherein the hair region image is configured to indicate a hair region of the target object in the original image, and the occluded face region image is configured to indicate an occluded face region of the target object in the original image; and
acquiring the second image by fusing the facial texture image with the first image comprises:
acquiring the second image based on the hair region image, the occluded face region image, the first image, the facial texture image, and the average brightness parameter.

10. The method according to claim 9 , wherein said acquiring the second image based on the hair region image, the occluded face region image, the first image, the facial texture image, and the average brightness parameter comprises:
acquiring a third image by multiplying a pixel value of a pixel in the hair region image by pixel values of pixels at corresponding positions in the occluded face region image, the first image, and the facial texture image; and
acquiring the second image by multiplying a pixel value of each pixel in the third image by the average brightness parameter.

11. An electronic device, comprising:
one or more processors; and
one or more memories configured to store one or more instructions executable by the one or more processors;
wherein the one or more processors, when loading and executing the one or more instructions, are caused to perform the following processes:
acquiring a first image by smoothing a skin region of a target object in an original image;
determining a skin texture material matching with a face area of the target object;
acquiring a facial texture image of the target object by rendering the skin texture material, wherein bo

Assignees

• Beijing Dajia Internet Information Tech Co Ltd

Inventors

• Guo Sainan
• Yang Di

Classifications

• G06T11/10 Primary

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

• G06T11/00

  Two-dimensional [2D] image generation · CPC title

• G06T2207/20221

  Image fusion; Image merging · CPC title

• G06T15/04

  Texture mapping · CPC title

• G06T2200/04

  involving 3D image data · CPC title