Although ray tracing is beginning to be used in real-time rendering, the time complexity of ray tracing is still high and the number of rays that can be emitted in real-time is limited. In this study, we propose affine-transformed ray alignment as a new method to reduce the time complexity of ray tracing.

Affine-Transformed Ray Alignment is a method to reduce the number of rays by combining multiple rays into a single ray using affine transformation. In addition, by making it possible to obtain the intersection information of multiple rays from a single ray after the ray reduction, the intersection information is the same as that obtained without the ray reduction. This has the advantage of achieving high speed and no errors in the rendered image.

Reduce the number of rays while maintaining relative intersection information


Name Affiliation Web site
Yuki Nishidate Keio University





Domestic presentations
  1. Yuki NishidateIssei Fujishiro: “Affine-Transformed Ray Alignment for Fast Ray Traversal,” in SIGGRAPH Asia 2022 Technical Communications (SA ’22 Technical Communications), December 06―09, 2022, Daegu, Korea, ACM, New York, NY, USA, Article 13, 4 pages, DOI: 10.1145/3550340.3564228
  2. Yuki NishidateIssei Fujishiro: “Affine-transformed ray alignment for fast ray traversal,” in proceedings of Visual Computing 2022, pp. 20:1―20:4, Kyoto, October 7, 2022, CGVI student presentation award (in Japanese).
  3. Yuki Nishidate, Issei Fujishiro: “Affine-Transformed Ray Alignment for Fast Primary Ray Traversal,” in Proceedings of the 84th National Convention of Information Proceedings Society of Japan, Vol. 4, pp. 247-248 (6ZF–01), Ehime University Johoku Campas, Ehime, March 3-5, 2022, Student encouragement award (in Japanese).
  4. Yuki Nishidate, Issei Fujishiro: “Affine-Transformed Ray Alignment for Accelerating Primary Ray Traversal – Implementation Using Embree -,” in Proceedings of SIG Technical Reports, Vol. 2022–CG–185, No. 2, pp. 1-6, March 11-12, 2022, Best presentation award (in Japanese).

Back to GP team page