Granted Patent B2

US 12,299,802 · App. 18/000,244 · Granted May 13, 2025

Inventors: Can Wang (Beijing, CN); Jiali Pan (Beijing, CN)

Assignee: BEIJING ZITIAO NETWORK TECHNOLOGY CO., LTD.

G06T15/04

Abstract

Claims (71)

1. An image processing method, comprising:

determining N texture images corresponding to a first image, and a correspondence between the N texture images and N model faces of a three-dimensional model, wherein N is an integer greater than 1;

determining, in the N texture images, a mapping region corresponding to each of the N model faces according to the correspondence and an offset coefficient, wherein the mapping region is an at least partial region of a texture image; and

mapping mapping regions in the N texture images to the three-dimensional model to obtain a three-dimensional image corresponding to the first image.

2. The method according to claim 1 , wherein the at least partial region is a middle region of the texture image; determining, in the N texture images, the mapping region corresponding to the each of the N model faces according to the correspondence and the offset coefficient comprises:

combining the N texture images into a region of L rows and K columns according to the correspondence to obtain a sampling region, wherein a product of L and K is N, and L and K are positive integers; and

determining the mapping region corresponding to the each of the N model faces according to the sampling region and the offset coefficient.

3. The method according to claim 2 , wherein for any one of the N model faces, determining the mapping region corresponding to the each of the N model faces according to the sampling region and the offset coefficient comprises:

determining a first position of the texture image corresponding to the any one of the N model faces in the sampling region;

acquiring vertex texture coordinates of the any one of the N model faces;

determining vertex sampling coordinates in the sampling region according to the vertex texture coordinates, the offset coefficient and the first position; and

determining the mapping region corresponding to the any one of the N model faces according to the vertex sampling coordinates and the first position.

4. The method according to claim 3 , wherein the vertex texture coordinates comprise vertex texture abscissas and vertex texture ordinates; determining the vertex sampling coordinates in the sampling region according to the vertex texture coordinates, the offset coefficient and the first position comprises:

determining abscissas of the vertex sampling coordinates according to the vertex texture abscissas, the first position, a number of columns of the sampling region and the offset coefficient;

determining ordinates of the vertex sampling coordinates according to the vertex texture ordinates, the first position, the number of columns of the sampling region, a number of rows of the sampling region and the offset coefficient; and

obtaining the vertex sampling coordinates according to the abscissas of the vertex sampling coordinates and the ordinates of the vertex sampling coordinates.

5. The method according to claim 3 , wherein a number of rows of the sampling region is 2; determining the mapping region corresponding to the any one of the N model faces according to the vertex sampling coordinates and the first position comprises:

upon determining that the first position is located in a first row of the sampling region, determining the mapping region corresponding to the any one of the N model faces according to the vertex sampling coordinates; and

upon determining that the first position is located in a second row of the sampling region, performing flipping processing of a preset angle on the vertex sampling coordinates to obtain target vertex sampling coordinates, and determining the mapping region corresponding to the any one of the N model faces according to the target vertex sampling coordinates.

6. The method according to claim 5 , wherein determining the mapping region corresponding to the any one of the N model faces according to the vertex sampling coordinates comprises:

determining a first region corresponding to the vertex sampling coordinates in the sampling region; and

determining the first region as the mapping region corresponding to the any one of the N model faces.

7. The method according to claim 5 , wherein determining the mapping region corresponding to the any one of the N model faces according to the target vertex sampling coordinates comprises:

determining a second region corresponding to the target vertex sampling coordinates in the sampling region; and

determining the second region as the mapping region corresponding to the any one of the N model faces.

8. The method according to claim 1 , before determining, in the N texture images, the mapping region corresponding to the each of the N model faces according to the correspondence and the offset coefficient, further comprising:

acquiring a preset region size of edges of the N texture images; and

determining the offset coefficient according to the preset region size.

9. The method according to claim 2 , before determining, in the N texture images, the mapping region corresponding to the each of the N model faces according to the correspondence and the offset coefficient, further comprising:

acquiring a preset region size of edges of the N texture images; and

determining the offset coefficient according to the preset region size.

10. The method according to claim 8 , wherein determining the offset coefficient according to the preset region size comprises:

acquiring a first preset relationship, wherein the first preset relationship comprises at least one region size and a coefficient corresponding to each region size; and

determining the offset coefficient according to the preset region size and the first preset relationship.

11. An image processing apparatus, comprising a memory and a processor;

the memory stores computer execution instructions;

the processor executes the computer execution instruction stored in the memory, so that the processor is configured to:

determine N texture images corresponding to a first image, and a correspondence between the N texture images and N model faces of a three-dimensional model, wherein N is an integer greater than 1;

determine, in the N texture images, a mapping region corresponding to each of the N model faces according to the correspondence and an offset coefficient, wherein the mapping region is an at least partial region of a texture image; and

map mapping regions in the N texture images to the three-dimensional model to obtain a three-dimensional image corresponding to the first image.

12. The apparatus according to claim 11 , wherein the at least partial region is a middle region of the texture image; the processor is configured to:

combine the N texture images into a region of L rows and K columns according to the correspondence to obtain a sampling region, wherein a product of L and K is N, and L and K are positive integers; and

determine the mapping region corresponding to the each of the N model faces according to the sampling region and the offset coefficient.

13. The apparatus according to claim 12 , wherein for any one of the N model faces, the processor is configured to:

determine a first position of the texture image corresponding to the any one of the N model faces in the sampling region;

acquire vertex texture coordinates of the any one of the N model faces;

determine vertex sampling coordinates in the sampling region according to the vertex texture coordinates, the offset coefficient and the first position; and

determine the mapping region corresponding to the any one of the N model faces according to the vertex sampling coordinates and the first position.

14. The apparatus according to claim 13 , wherein the vertex texture coordinates comprise vertex texture abscissas and vertex texture ordinates; the processor is configured to:

determine abscissas of the vertex sampling coordinates according to the vertex texture abscissas, the first position, a number of columns of the sampling region and the offset coefficient;

determine ordinates of the vertex sampling coordinates according to the vertex texture ordinates, the first position, the number of columns of the sampling region, a number of rows of the sampling region and the offset coefficient; and

obtain the vertex sampling coordinates according to the abscissas of the vertex sampling coordinates and the ordinates of the vertex sampling coordinates.

15. The apparatus according to claim 13 , wherein a number of rows of the sampling region is 2; the processor is configured to:

upon determining that the first position is located in a first row of the sampling region, determine the mapping region corresponding to the any one of the N model faces according to the vertex sampling coordinates; and

upon determining that the first position is located in a second row of the sampling region, perform flipping processing of a preset angle on the vertex sampling coordinates to obtain target vertex sampling coordinates, and determine the mapping region corresponding to the any one of the N model faces according to the target vertex sampling coordinates.

16. The apparatus according to claim 15 , wherein the processor is configured to:

determine a first region corresponding to the vertex sampling coordinates in the sampling region; and

determine the first region as the mapping region corresponding to the any one of the N model faces.

17. The apparatus according to claim 15 , wherein the processor is configured to:

determine a second region corresponding to the target vertex sampling coordinates in the sampling region; and

determine the second region as the mapping region corresponding to the any one of the N model faces.

18. The apparatus according to claim 11 , wherein the processor is further configured to:

acquire a preset region size of edges of the N texture images; and

determine the offset coefficient according to the preset region size.

19. The apparatus according to claim 18 , wherein the processor is configured to:

acquire a first preset relationship, wherein the first preset relationship comprises at least one region size and a coefficient corresponding to each region size; and

determine the offset coefficient according to the preset region size and the first preset relationship.

20. A non-transitory computer-readable storage medium having computer execution instructions stored therein, wherein when the computer execution instructions are executed by a processer, the following operations are implemented:

determining N texture images corresponding to a first image, and a correspondence between the N texture images and N model faces of a three-dimensional model, wherein N is an integer greater than 1;

determining, in the N texture images, a mapping region corresponding to each of the N model faces according to the correspondence and an offset coefficient, wherein the mapping region is an at least partial region of a texture image; and

mapping mapping regions in the N texture images to the three-dimensional model to obtain a three-dimensional image corresponding to the first image.

Assignments (3)

Priority Claims (1)

Continuity (1)

References Cited (31)