Description:
This project focused on optimising the real-time rendering of over 250,000 stars from the Gaia DR2 and RAVE DR5 datasets for a low-power computer telescope overlay. To overcome hardware limitations, I implemented spatial indexing using an Octree data structure to perform efficient frustum culling, reducing unnecessary rendering calculations. The solution achieved a significant performance improvement over traditional methods by cutting processing complexity from O(n) to O(log n + k). Through multithreading, memory optimisation, and benchmarking with Google Bench, the system was refined for responsiveness and scalability, demonstrating my ability to design efficient algorithms and optimise performance for constrained embedded environments.
Links: