Project Abstract

In the rapidly evolving field of computer graphics, the demand for real-time rendering capabilities is increasingly urgent, particularly in applications such as virtual reality and advanced gaming systems. This project investigates the integration of Kd-trees into ray tracing algorithms to enhance computational efficiency and enable the rendering of complex scenes in real time without sacrificing image quality. By employing Kd-trees, a well-established spatial data structure, this research aims to reduce the number of ray-object intersection tests, a notorious bottleneck in traditional ray tracing processes. The methodological approach includes the implementation and testing of Kd-trees within a Java-based ray tracing framework, comparing performance metrics such as rendering time and computational overhead against traditional methods. Preliminary simulations suggest that Kd-trees can significantly expedite the ray tracing process by efficiently managing spatial queries and minimizing unnecessary calculations. The expected outcome of this research is a demonstrable improvement in rendering speeds and a reduction in computational demands, facilitating more sophisticated visualizations in real-time applications. If successful, this project will contribute to the graphics field by providing a robust solution that enhances the capabilities of ray tracing technology, pushing the boundaries of what is possible in digital visualization and interactive media.

Keywords: Ray Tracing, Kd-Trees, 3D Rendering

 Conference Details

Session: Presentation Stream 3 at Presentation Slot 8

Location: GH029 at Tuesday 7th 13:30 – 17:00

Markers: Deb Roy, Solmaz safari

Course: MSc Computer Science, Masters PG

Future Plans: I’m looking for work