Achieving robust autonomy in robotics is challenged by sensor data reliability, especially in extreme environments. Conventional sensors like RGB cameras fail in low-visibility conditions (e.g., smoke, fog, low light), while active sensors such as Radar and Lidar struggle with reflective medium and particle obscurants and often lack dense information at close range. Thermal infrared cameras offer advantages for robust sensing: they work well in low light, provide high-resolution imaging, and are crucial for search and rescue and hazard detection. However, they face challenges, such as noise from flat field correction, ambiguous tone-mapping, difficult calibration, and temperature drifts. With the advent of compact, affordable thermal cameras, interest in thermal imaging for robotics is growing beyond military and surveillance usage.
This workshop aims to advance the emerging field by providing a platform for sharing developments in localization, mapping, navigation, segmentation, detection, and other methods tailored for thermal infrared imagery. We will invite leading researchers and experts to share insights and breakthroughs while encouraging contributions from newcomers and diverse backgrounds. Attendees will engage in discussions, network with experts, and explore future directions, making it invaluable for advancing thermal imaging in robotics.
