# US 2024/0203027 A1 · Device and method for generating dynamic texture map for 3 dimensional digital human

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2024/0203027 A1
- Document ID: US20240203027A1
- Status: Published application
- Kind: A1
- Application number: 18/544,111
- Filed: 2023-12-18
- Published: 2024-06-20
- Earliest claimed: 2022-12-20
- Assignees: Electronics and Telecommunications Research Institute
- Applicants: Electronics and Telecommunications Research Institute
- Inventors: Seung Uk YOON; Ki Nam KIM; Tae Joon KIM; Seung Wook LEE; Bon Woo HWANG
- Web page: https://patentgrep.com/patent/US20240203027A1

## CPC

- G06T 15/04: Texture mapping
- G06T 5/50: using two or more images, e.g. averaging or subtraction
- G06T 5/77: Retouching; Inpainting; Scratch removal
- G06T 17/00: Three-dimensional [3D] modelling for computer graphics
- G06T 2207/20021: Dividing image into blocks, subimages or windows
- G06T 2207/30196: Human being; Person
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

A device and a method for generating dynamic texture map for 3 dimensional digital human are disclosed. The method comprises generating high-quality images from low-quality images including a low-quality object; generating partial high-quality images including a high-quality object by replacing the low-quality object within the low-quality images with the high-quality object within the high-quality images; and generating a high-quality texture map for a 3D model corresponding to the high-quality object using the partial high-quality images.

## Claims

### Claim 1

A computer-implemented method for generating a high-quality texture map for a digital human comprising:
generating high-quality images from low-quality images including a low-quality object;
generating partial high-quality images including a high-quality object by replacing the low-quality object within the low-quality images with the high-quality object within the high-quality images; and
generating a high-quality texture map for a 3D model corresponding to the high-quality object using the partial high-quality images.

### Claim 2 (depends on 1)

The method of claim 1, wherein the low-quality images indicate key frame images among rendered images.

### Claim 3 (depends on 1)

The method of claim 1, wherein the generating of the partial high-quality images includes:
separating the low-quality object from the background of the low-quality images;
separating the high-quality object from the background of the high-quality images; and
replacing at least a part of the low-quality object with a corresponding part of the high-quality object.

### Claim 4 (depends on 1)

The method of claim 1, wherein the generating of the high-quality texture map for the 3D model includes:
generating the 3D model with the same shape and pose as the high-quality object within the partial high-quality images and with a partial texture map related to the high-quality object;
separating the 3D model into the shape and the partial texture map; and
generating a complete texture map by performing inpainting on the partial texture map.

### Claim 5 (depends on 1)

The method of claim 1, wherein the low-quality images are obtained by rendering the shape of the 3D model and the low-quality texture map for the 3D model.

### Claim 6 (depends on 5)

The method of claim 5, wherein the generating of the high-quality texture map for the 3D model includes:
generating the high-quality texture map by applying the high-quality object to the low-quality texture map based on a mapping relationship between each part of the low-quality texture map and each part of the high-quality object.

### Claim 7

A device for generating a high-quality texture map for a digital human comprising:
at least one memory storing computer-executable instructions; and
at least one processor,
wherein the at least one processor is configured to execute the computer-executable instructions to:
generates high-quality images from low-quality images including low-quality objects;
generates partial high-quality images including a high-quality object by the low-quality object within the low-quality images with a high-quality object within the high-quality images; and
generates a high-quality texture map for a 3D model corresponding to the high-quality object using the partial high-quality images.

### Claim 8 (depends on 7)

The device of claim 7, wherein the low-quality images indicate key frame images among rendered images.

### Claim 9 (depends on 7)

The device of claim 7, wherein the at least one processor:
separates the low-quality object from the background of the low-quality images;
separates the high-quality object from the background of the high-quality images; and
replaces at least a part of the low-quality object with a corresponding part of the high-quality object.

### Claim 10 (depends on 7)

The device of claim 7, wherein the at least one processor:
generates the 3D model with the same shape and pose as the high-quality object within the partial high-quality images and with a partial texture map related to the high-quality object;
separates the 3D model into the shape and the partial texture map; and
generates a complete texture map by performing inpainting on the partial texture map.

### Claim 11 (depends on 7)

The device of claim 7, wherein the low-quality images are obtained by rendering the shape of the 3D model and the low-quality texture map for the 3D model.

### Claim 12 (depends on 11)

The device of claim 11, wherein the at least one processor:
generates the high-quality texture map by applying the high-quality object to the low-quality texture map based on a mapping relationship between each part of the low-quality texture map and each part of the high-quality object.
