Granted Patent B2

US 12,299,817 · App. 17/985,102 · Granted May 13, 2025

Inventor: Zili Ding (Shenzhen, CN)

Assignee: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITED

G06T17/05A63F13/63G06T19/20G06T2210/56

Abstract

Claims (97)

1. A virtual scene generation method, performed by a computer device, the method comprising:

acquiring three-dimensional point cloud space data of a real scene in real world and scene distribution data corresponding to the real scene, wherein the scene distribution data comprises point set data that indicates position and size information of scene objects in the real scene and the scene distribution data is distinct from the three-dimensional point cloud space data;

performing alignment on a scene area corresponding to the scene distribution data and a scene area corresponding to the three-dimensional point cloud space data to obtain a scene aligned area;

selecting the scene distribution data matching the scene aligned area from the scene distribution data to obtain aligned scene distribution data, further comprising:

selecting complete point set data from the scene distribution data when the scene aligned area has complete point set data of a first object type of scene objects;

eliminating the complete point set data from the scene distribution data when the scene aligned area has incomplete point set data of the first object type of scene objects;

selecting point set data of a second object type of scene objects in the scene aligned area from the scene distribution data;

extracting terrain features from the three-dimensional point cloud space data;

generating a three-dimensional terrain model of the real scene according to the terrain features;

extracting, from the aligned scene distribution data, scene object features of the scene objects in the real scene based on the position and size information of the scene objects in the real scene; and

generating scene object models corresponding to the scene object features in the three-dimensional terrain model to obtain a virtual scene corresponding to the real scene.

2. The method according to claim 1 , wherein:

the three-dimensional space data comprises terrain space data; and

extracting terrain features from the three-dimensional space data comprises:

determining the terrain features according to height differences between adjacent terrain points in the terrain space data.

3. The method according to claim 2 , wherein generating the three-dimensional terrain model according to the terrain features comprises:

determining terrain types corresponding to terrain areas in the terrain space data according to the terrain features;

adding corresponding terrain surface attribute information to the terrain areas according to the terrain types; and

generating the three-dimensional terrain model according to the terrain space data and the added terrain surface attribute information.

4. The method according to claim 1 , wherein:

the three-dimensional space data comprises cover object space data; and

the method further comprises:

determining a terrain surface coverage area in the three-dimensional terrain model according to the cover object space data; and

generating terrain surface cover objects in the terrain surface coverage area.

5. The method according to claim 4 , wherein determining the terrain surface coverage area comprises:

determining terrain points corresponding to terrain surface cover points in the cover object space data in the three-dimensional terrain model;

selecting target terrain surface cover points according to height differences between the terrain surface cover points and the corresponding terrain points;

determining overlapping areas of expansion areas and the three-dimensional terrain model, the expansion areas being areas obtained after expansion based on the target terrain surface cover points; and

determining the terrain surface coverage area in the three-dimensional terrain model according to the overlapping areas.

6. The method according to claim 4 , wherein generating the terrain surface cover objects in the terrain surface coverage area comprises:

determining generation parameter information of the terrain surface cover objects in the terrain surface coverage area; and

generating the terrain surface cover objects in the terrain surface coverage area according to the generation parameter information.

7. The method according to claim 1 , wherein extracting the scene object features comprises:

determining positions and heights of the scene objects in the real scene in the three-dimensional terrain model based on the scene distribution data;

generating three-dimensional scene object data corresponding to the scene objects according to the positions and the heights; and

extracting the scene object features of the scene objects from the three-dimensional scene object data.

8. The method according to claim 7 , wherein determining the positions and heights of the scene objects in the real scene comprises:

determining the positions and the heights of the scene objects in the three-dimensional terrain model according to original positions when the scene distribution data includes the original positions of the scene objects in the real scene.

9. The method according to claim 1 , wherein generating the scene object models comprises:

generating initial scene object models corresponding to the scene objects according to the scene object features;

adding scene object attribute information conforming to object types to the initial scene object models according to the object types of the scene objects to obtain the scene object models matched with the scene objects; and

adding the scene object models to the three-dimensional terrain model to obtain the virtual scene corresponding to the real scene.

10. The method according to claim 9 , wherein adding the scene object models to the three-dimensional terrain model to obtain the virtual scene corresponding to the real scene comprises:

determining corresponding terrain areas of the scene object models in the three-dimensional terrain model;

performing smoothing treatment on the terrain areas to obtain smoothened terrain areas; and

adding the scene object models to the smoothened terrain areas in the three-dimensional terrain model to obtain the virtual scene corresponding to the real scene.

11. The method according to claim 10 , wherein determining the corresponding terrain areas of the scene object models in the three-dimensional terrain model comprises:

determining a position area corresponding to the scene object models in the three-dimensional terrain model and an adjacent area of the position area as the terrain areas corresponding to the scene object models.

12. A computer device, comprising:

one or more processors; and

memory storing one or more programs, the one or more programs comprising instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising:

acquiring three-dimensional point cloud space data of a real scene in real world and scene distribution data corresponding to the real scene, wherein the scene distribution data comprises point set data that indicates position and size information of scene objects in the real scene and the scene distribution data is distinct from the three-dimensional point cloud space data;

performing alignment on a scene area corresponding to the scene distribution data and a scene area corresponding to the three-dimensional point cloud space data to obtain a scene aligned area;

selecting the scene distribution data matching the scene aligned area from the scene distribution data to obtain aligned scene distribution data, further comprising:

selecting complete point set data from the scene distribution data when the scene aligned area has complete point set data of a first object type of scene objects;

eliminating the complete point set data from the scene distribution data when the scene aligned area has incomplete point set data of the first object type of scene objects;

selecting point set data of a second object type of scene objects in the scene aligned area from the scene distribution data;

extracting terrain features from the three-dimensional point cloud space data;

generating a three-dimensional terrain model of the real scene according to the terrain features;

extracting, from the aligned scene distribution data, scene object features of the scene objects in the real scene based on the position and size information of the scene objects in the real scene; and

generating scene object models corresponding to the scene object features in the three-dimensional terrain model to obtain a virtual scene corresponding to the real scene.

13. The computer device according to claim 12 , wherein:

the three-dimensional space data comprises terrain space data; and

extracting terrain features from the three-dimensional space data comprises:

determining the terrain features according to height differences between adjacent terrain points in the terrain space data.

14. The computer device according to claim 12 , wherein:

the three-dimensional space data comprises cover object space data; and

the operations further comprise:

determining a terrain surface coverage area in the three-dimensional terrain model according to the cover object space data; and

generating terrain surface cover objects in the terrain surface coverage area.

15. The computer device according to claim 12 , wherein extracting the scene object features comprises:

determining positions and heights of the scene objects in the real scene in the three-dimensional terrain model based on the scene distribution data;

generating three-dimensional scene object data corresponding to the scene objects according to the positions and the heights; and

extracting the scene object features of the scene objects from the three-dimensional scene object data.

16. The computer device according to claim 12 , wherein generating the scene object models comprises:

generating initial scene object models corresponding to the scene objects according to the scene object features;

adding scene object attribute information conforming to object types to the initial scene object models according to the object types of the scene objects to obtain the scene object models matched with the scene objects; and

adding the scene object models to the three-dimensional terrain model to obtain the virtual scene corresponding to the real scene.

17. A non-transitory computer-readable storage medium, storing one or more instructions, the one or more instructions, when executed by one or more processors of a computer device, cause the computer device to perform operations comprising:

acquiring three-dimensional point cloud space data of a real scene in real world and scene distribution data corresponding to the real scene, wherein the scene distribution data comprises point set data that indicates position and size information of scene objects in the real scene and the scene distribution data is distinct from the three-dimensional point cloud space data;

performing alignment on a scene area corresponding to the scene distribution data and a scene area corresponding to the three-dimensional point cloud space data to obtain a scene aligned area;

selecting the scene distribution data matching the scene aligned area from the scene distribution data to obtain aligned scene distribution data, further comprising:

selecting complete point set data from the scene distribution data when the scene aligned area has complete point set data of a first object type of scene objects;

eliminating the complete point set data from the scene distribution data when the scene aligned area has incomplete point set data of the first object type of scene objects;

selecting point set data of a second object type of scene objects in the scene aligned area from the scene distribution data;

extracting terrain features from the three-dimensional point cloud space data;

generating a three-dimensional terrain model of the real scene according to the terrain features;

extracting, from the aligned scene distribution data, scene object features of the scene objects in the real scene based on the position and size information of the scene objects in the real scene; and

generating scene object models corresponding to the scene object features in the three-dimensional terrain model to obtain a virtual scene corresponding to the real scene.

18. The non-transitory computer-readable storage medium according to claim 17 , wherein generating the scene object models comprises:

generating initial scene object models corresponding to the scene objects according to the scene object features;

adding scene object attribute information conforming to object types to the initial scene object models according to the object types of the scene objects to obtain the scene object models matched with the scene objects; and

adding the scene object models to the three-dimensional terrain model to obtain the virtual scene corresponding to the real scene.

19. The non-transitory computer-readable storage medium according to claim 17 , wherein extracting the scene object features comprises:

determining positions and heights of the scene objects in the real scene in the three-dimensional terrain model based on the scene distribution data;

generating three-dimensional scene object data corresponding to the scene objects according to the positions and the heights; and

extracting the scene object features of the scene objects from the three-dimensional scene object data.

Assignments (1)

Priority Claims (1)

Continuity (2)

References Cited (28)