Tracking the Invisible: Self-Supervised Wheel Track Detection with Thermal-RGB Fusion
What it is
This paper proposes a self-supervised neural network method for detecting wheel tracks on snow-covered roads using synchronized RGB and Long-Wave Infrared (LWIR) imagery. It automatically generates training labels by projecting future wheel trajectories from recorded GNSS data, eliminating the need for manual annotation.
The problem it solves
Standard traversability estimation methods fail on snow-covered roads where visual features are monotonous and existing wheel tracks are critical for safety but difficult to detect.
What you could use it for
- Autonomous driving in winter conditions
- Advanced driver assistance systems (ADAS) for snow-covered roads
- Path planning for vehicles in white-out environments
Read the source ↗Back to the library →
Summary written by an AI model from the source; check the source before you rely on it.