Granted Patent B2

US 12,315,078 · App. 18/005,940 · Granted May 27, 2025

Inventors: Yu Zhang (Jiangsu, CN); Zhan Gong (Jiangsu, CN)

Assignee: INSPUR SUZHOU INTELLIGENT TECHNOLOGY CO., LTD.

G06T17/005G06T15/005G06T2210/22

Abstract

Claims (71)

1. A self-driving simulation rendering method, comprising:

loading a scene file, performing scene modeling based on a tree model, and establishing a static spatial index;

setting a dynamic traffic flow to generate a dynamic scene, and setting a position and direction of a sensor; and

creating a first thread and a second thread, scheduling the first thread to crop each frame of the dynamic scene based on the tree model, and scheduling the second thread to render, based on the position and direction of the sensor, each frame cropped by the first thread and output the rendered frame;

wherein scheduling the first thread to crop each frame of the dynamic scene based on the tree model comprises:

determining whether an area, corresponding to the tree model, of the dynamic scene is a rendering area of interest,

further determining a relationship between the rendering area of interest and a cropping area in a case that the area, corresponding to the tree model, of the dynamic scene is the rendering area of interest; and

constructing a Three-dimensional (3D) scene and outputting a cropping result in a case that the rendering area of interest is in the cropping area;

wherein the self-driving simulation rendering method further comprising:

cropping and filtering the tree model in a case that the rendering area of interest intersects the cropping area; and

reconstructing a 3D scene model and outputting a cropping result.

2. The self-driving simulation rendering method according to claim 1 , wherein in a case where there are multiple sensors, creating the first thread and the second thread comprises:

creating corresponding viewports according to the number of the multiple sensors, and respectively creating a first thread and a second thread for each viewport; and

scheduling the first threads of all viewports to start cropping a first frame of the dynamic scene at the same time.

3. The self-driving simulation rendering method according to claim 1 , wherein scheduling the first thread to crop each frame of the dynamic scene based on the tree model comprises:

cropping, by the first thread, a current frame of the dynamic scene based on the tree model; and

cropping a next frame in a case that the cropping of the current frame is completed.

4. The self-driving simulation rendering method according to claim 1 , wherein scheduling the second thread to render, based on the position and direction of the sensor, each frame cropped by the first thread and output the rendered frame comprises:

rendering, by the second thread, a Three-dimensional (3D) scene obtained by cropping a current frame by the first thread into a Two-dimensional (2D) picture based on the position and direction of the sensor, and outputting, by the second thread, the 2D picture;

determining, in a case that the rendering of the current frame is completed, whether the first thread completes the cropping of a next frame; and

scheduling the second thread to render and output the next frame in a case that the first thread completes the cropping of the next frame.

5. The self-driving simulation rendering method according to claim 4 , wherein rendering, by the second thread, the 3D scene obtained by cropping the current frame by the first thread into the 2D picture according to the position and direction of the sensor comprises:

render, by the second thread, the 3D scene into the 2D picture for outputting by using an Open Graphics Library (OpenGL) according to the position and direction of the sensor.

6. The self-driving simulation rendering method according to claim 1 , wherein establishing the static spatial index comprises:

establishing an east primary node, a west primary node, a south primary node, and a north primary node according to directions of the scene; and

establishing one or more secondary nodes according to objects in the scene.

7. The self-driving simulation method according to claim 1 , wherein the first thread and the second thread operate on different Central Processing Unit (CPU) cores.

8. The self-driving simulation rendering method according to claim 1 , wherein cropping and filtering the tree model comprises:

determining, according to a binary tree model, scenes and vehicles needing to be included in a part to be rendered;

acquiring an area range corresponding to the sensor according to the area range, and then corresponding the objects needed to be rendered to object information included in the binary tree model.

9. The self-driving simulation rendering method according to claim 8 , further comprising:

cropping the scene based on the tree model, wherein only a cropping area needs to be processed based on the tree model, and data outside the cropping area is directly excluded.

10. The self-driving simulation rendering method according to claim 9 , further comprising:

respectively filtering and cropping the tree model stage by stage, so as to obtain a cropping result;

reconstructing the cropping result according to the tree model, so as to obtain a scene in the cropping area.

11. The self-driving simulation rendering method according to claim 1 , wherein establishing a static spatial index comprises:

establishing the static spatial index by taking the tree model as a basis.

12. A computer device, comprising:

at least one processor; and

a memory, wherein the memory stores a computer instruction capable of running on the at least one processor; and the instruction, when being executed by the at least one processor, causes the at least one processor to:

load a scene file, perform scene modeling based on a tree model, and establish a static spatial index;

set a dynamic traffic flow to generate a dynamic scene, and set a position and direction of a sensor; and

create a first thread and a second thread, schedule the first thread to crop each frame of the dynamic scene based on the tree model, and schedule the second thread to render, based on the position and direction of the sensor, each frame cropped by the first thread and output the rendered frame;

wherein the instruction, when being executed by the at least one processor, causes the at least one processor to schedule the first thread to crop each frame of the dynamic scene based on the tree model, by following steps:

determining whether an area, corresponding to the tree model, of the dynamic scene is a rendering area of interest;

further determining a relationship between the rendering area of interest and a cropping area in a case that the area, corresponding to the tree model, of the dynamic scene is the rendering area of interest; and

constructing a Three-dimensional (3D) scene and outputting a cropping result in a case that the rendering area of interest is in the cropping area;

wherein the instruction, when being executed by the at least one processor, further causes the at least one processor to:

crop and filter the tree model in a case that the rendering area of interest intersects the cropping area; and

reconstruct a 3D scene model and output a cropping result.

13. The computer device according to claim 12 , wherein the instruction, when being executed by the at least one process, causes the at least one processor to:

create corresponding viewports according to the number of the multiple sensors, and respectively creating a first thread and a second thread for each viewport; and

schedule the first threads of all viewports to start cropping the first frame of the dynamic scene at the same time.

14. The computer device according to claim 12 , wherein the instruction, when being executed by the at least one processor, causes the at least one processor to:

crop, by the first thread, a current frame of the dynamic scene based on the tree model; and

crop a next frame in a case that the cropping of the current frame is completed.

15. The computer device according to claim 12 , wherein the instruction, when being executed by the at least one processor, causes the at least one processor to:

render, by the second thread, a Three-dimensional (3D) scene obtained by cropping a current frame by the first thread into a Two-dimensional (2D) picture based on the position and direction of the sensor, and outputting by the second thread, the 2D picture;

determine, in a case that the rendering of the current frame is completed, whether the first thread completes the cropping of a next frame; and

schedule the second thread to render and output the next frame in a case that the first thread completes the cropping of the next frame.

16. A non-transitory computer-readable storage medium, storing a computer program, wherein when the computer program is executed by at least one processor, following operations are implemented:

loading a scene file, performing scene modeling based on a tree model, and establishing a static spatial index;

a position and direction of a sensor: and setting a dynamic traffic flow to generate a dynamic scene, and setting

creating a first thread and a second thread, scheduling the first thread to crop each frame of the dynamic scene based on the tree model, and scheduling the second thread to render, based on the position and direction of the sensor, each frame cropped by the first thread and output the rendered frame;

wherein when the computer program is executed by at least one processor, following operations are implemented to schedule the first thread to crop each frame of the dynamic scene based on the tree model;

determining whether an area, corresponding to the tree model, of the dynamic scene is a rendering area of interest;

further determining a relationship between the rendering area of interest and a cropping area in a case that the area, corresponding to the tree model, of the dynamic scene is the rendering area of interest; and

constructing a Three-dimensional (3D) scene and outputting a cropping result in a case that the rendering area of interest is in the cropping area;

wherein when the computer program is executed by at least one processor, following operations are further implemented;

cropping and filtering the tree model in a case that the rendering area of interest intersects the cropping area, and

reconstructing a 3D scene model and outputting a cropping result.

Assignments (2)

Priority Claims (1)

Continuity (1)

References Cited (16)