Acceleration of shooting and bouncing ray method based on OptiX and normal vectors correction.

• Main Authors: Gaogui Xu, Chunzhu Dong, Tao Zhao, Hongcheng Yin, Xuan Chen
• Format: Article
• Language: English
• Published: Public Library of Science (PLoS) 2021-01-01
• Series: PLoS ONE
• Online Access: https://doi.org/10.1371/journal.pone.0253743

The present paper deals with a new efficient shooting and bouncing ray (SBR) method based on OptiX and normal vectors correction. The basic idea is to make full use of the computing resources of the RTX series graphics cards. For ray tracing, the algorithm uses OptiX to invoke the built-in RT Cores of hardware. Thus, a fast intersection test can be implemented. To reduce the error of ray tracing caused by the facetted surface characterizing the curved surface, the direction of the reflected ray is corrected by normal vectors correction. Additionally, multiple GPU cores are invoked to accelerate the calculation of far-field integration of millions of ray tubes, which can improve the efficiency of the algorithm while reducing the data transmission time of heterogeneous devices. Simulation results show that the ray path after normal vectors correction is consistent with the theoretical results, and the algorithm can predict the RCS of arbitrary facetted geometries, which is 60 times faster than the SBR method based on kd-tree.