Light parallel intersection method based on many-core processor and optical path tracking system

A technology of many-core processors and rays, which is applied in the field of ray parallel intersecting method and optical path tracing system, can solve the problems of consuming calculation time and reducing the efficiency of optical path tracing algorithm, so as to improve computing efficiency, shorten optical path tracing time, and ensure rendering quality effect

CN111402388AInactive Publication Date: 2020-07-10SHANDONG UNIV
5 Cites 1 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Publication Date
2020-07-10
Estimated Expiration
Not applicable · inactive patent

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention discloses a light parallel intersection method based on a many-core processor and an optical path tracking system. The method comprises the steps: opening a global data area used for storing sampling light rays in a main core, performing light ray parallel intersection on the sampling light rays from a core, storing intersection results in the global data area, performing coloring operation by the main core according to the intersection results, and calculating the integral value of the sampling light rays on a light path; separating light intersection and coloring operations from each other; calculating a time ratio of light intersection to coloring operation; dividing the task load of the two; synchronously carrying out the second part of light intersection executed in theslave core and the first part of coloring operation executed in the master core, so the memory of the slave core is fully utilized, the intersection efficiency is effectively improved by utilizing a parallel light intersection mode, the total time of an optical path tracking algorithm is reduced, and the intersection accuracy can be ensured, thereby guaranteeing the drawing quality of an image.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The present disclosure relates to the technical field of graphic real rendering, and in particular to a method for intersecting rays in parallel based on many-core processors and an optical path tracing system. Background technique

[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art.

[0003] Optical path tracing algorithm is a physically-based realistic rendering method, which uses the physical principles of the real world to simulate the visual sense generated by the interaction of light and matter in the virtual world. The goal of photorealistic rendering is to make images generated from 3D scenes of the same scene indistinguishable from real-world photographs. The optical path tracing algorithm is based on the perspective of the observer, according to the reversible physical principle of light, reversely traces the light, emits virtual light fro...

Examples

Embodiment 1

[0034] The light path tracing algorithm process is to launch a ray from the camera or observer’s point of view, and trace the ray in reverse according to the reversible optical principle of light. The ray first intersects with the scene, and then performs the coloring operation at the intersection point to generate reflected or refracted rays, namely The next ray (secondary ray) is traced until the ray intersects the light source. This is a complete optical path of a ray, and the integration is performed on this optical path to calculate the contribution value of this ray to the light radiance of a pixel.

[0035] In this embodiment, a method for intersecting rays in parallel based on many-core processors is provided, such as figure 1 shown, including:

[0036] Open up a global data area for storing sampling rays in the main core, perform ray intersection of the sampling rays in parallel from the checker, and store the intersection result in the global data area, and the main...

Embodiment 2

[0069] Such as Figure 4 As shown, this embodiment provides a parallel intersection method for ray occlusion judgment based on many-core processors, which executes the occlusion judgment after the intersection point samples the light source in parallel during the shading process, reduces the time consumed in the shading process, and effectively shortens the optical path. Tracked computation time, including:

[0070] Opening up a global data area for storing sampling rays in the main core, performing parallel intersecting of the sampling rays from the checker, and storing the intersecting results in the global data area; the main core performs coloring operations according to the intersecting results;

[0071] In the coloring operation, the light source is sampled according to the intersection point to obtain the shadow sampling light, and the shadow sampling light is checked in parallel to determine whether the shadow sampling light is blocked, and the main core updates the in...

Embodiment 3

[0091] Such as Figure 5 As shown, this embodiment provides an optical path tracing system based on many-core processors, including several core groups, the core groups include master cores and slave cores, and the optical path tracing is divided in the master core of any core group , assign the divided light path tracing subtasks to other core groups;

[0092] The core group completes the steps described in the parallel intersecting method of rays based on many-core processors or the parallel intersecting method of ray occlusion judgment based on many-core processors.

[0093] The process of calling other core groups in the full-chip sharing mode is: initialize the runtime parameters of each core group, and distinguish the memory layout in each shared area to prevent concurrent reads and writes between core groups and the occurrence of dirty data. question;

[0094] Use the pthread library to initialize other core groups before running, and then hand over the executed work ...