Rift Rally

Official website: https://www.riftrally.com

During my time at Velan Studios as a game programmer, a focus of mine was computer vision research for the Rift Rally project. This innovative game integrated real-world RC cars with augmented reality elements, utilizing the car's onboard camera to overlay virtual content onto the live video feed.

The core objectives of my research for the game was to calibrate the car's camera to align with the virtual in-game camera, as well as assessing the robustness of object detection across various environmental conditions. This endeavor entailed an extensive data collection effort, involving the acquisition of a substantial volume of images captured by the car’s camera looking at calibration checkboards, object detection targets, and known false positive targets.

I authored numerous Python scripts to meticulously process and evaluate the image data. To gain a comprehensive understanding of the results, I leveraged the matplotlib library to create data visualizations. These visualizations were instrumental in evaluating the effectiveness of the object detection techniques in diverse scenarios.

Ultimately, I translated my findings into action by implementing object detection functionality within our custom game engine, employing the OpenCV library.

WORK IN PROGRESS...