Accurate Tracking in Motion – Real-Time Image Stability
The goal was to achieve flawless image tracking in motion, ensuring that the augmented reality experience remains stable even under challenging conditions. The focus was on eliminating issues such as flickering or vanishing augmentations, especially when the user moves or holds the marker at skewed or narrow angles. We used image recognition algorithms designed to keep augmentations seamlessly in place, even when the marker—such as a card—was barely visible. This solution resulted in a natural, immersive AR experience where the augmentation remains firmly anchored in reality, responding dynamically to hand movements and varying angles without disruptions.
Drift-Resistant Augmentation – Stability in Low-Minimal Drift in Poor SLAM Conditions
This challenge arose when working with a restaurant environment where low light conditions caused instability in the augmentation due to weak SLAM (Simultaneous Localization and Mapping). The augmentations would frequently drift or lose their anchor, resulting in a poor user experience. To address this, we developed a solution that ensured the augmentations remained stable, even in dark or low-light settings, with minimal drift. This enhancement significantly improved the user experience by maintaining consistent and reliable AR interactions, regardless of environmental conditions.
Real-Time Image Tracking – Low-Light Test with Unlit Marker
This challenge involved ensuring reliable image tracking in low-light conditions without relying on backlit markers or external lighting. The objective was to maintain consistent tracking on an unlit paper marker, even when it was folded, twisted, or not fully visible. By engineering a robust solution, we achieved a 70% confidence level in real-time image tracking, ensuring smooth and accurate detection even in minimal light, which significantly improved the AR experience in challenging environments.