| Ray tracing is a widely used algorithm that produced high-quality computer generated images and real global illumination. This is the traditional raster-based imaging method can not match. Ray tracing is the main idea to produce image in the future. However, the algorithm requires enormous amounts of computation, and as a result cannot be efficiently done in real-time on commodity CPU hardware.Therefore, research on accelerating ray tracing becomes necessary. In particular, with the development of general-purpose computation on Graphics Process Unit and scene segmentation technique, It provides acceleration of ray tracing with a good soil.In this paper, there are two aspects to accelerate ray tracing: software and hardware. With software accelerations, Scene-splitting technology is put to use to split the scene reasonably that can reduce the amount of calculations between rays and polygons in the scene. GPU hardware acceleration is put to use in parallel processor capacity, ray tracing calculations as the color value of each pixel is an independent process, it can create a lot of threads to simultaneously calculate multiple pixel.In addition, this paper designs and implements a ray tracing acceleration system. FBX scene file is used as the carriers of the scenes of this system. The system can extract necessary information from FBX file. Bounding volume hierarchies technique is adopted to split the loaded scene. The core part of ray tracing computation will be modified to adapt the advantages and limitations of CUDA so as to amplify the power of parallelization.This paper is divided into four parts. The first part describes the research status to accelerate ray tracing; The second part is the necessary research on principle and algorithm of ray tracing. The third part is the design and implementation of my ray tracing acceleration system. The fourth part is test and performance of my system with analysis, summarizes my work and the future anticipation of my study. |
PDF Full Text Request