BAMBI - Biodiversity Airborne Monitoring Based on Intelligent UAV sampling

In the light of the impending climate catastrophe as well as human encroachment into natural habitats (forests, seas, meadows, etc.), the integrity of these habitats is increasingly endangered – both worldwide as well as in Austria. Balanced ecosystems such as forests are the foundation of life on our planet, along with the underlying complex structures of diverse species of animals, plants, microbes, and other organisms living in symbiosis with each other. The preservation and promotion of biodiversity are central to maintaining this balance.

Counteracting the loss of biodiversity and habitats (both due to the extinction of species, but also due to overpopulation and associated negative impact on forest ecosystems) is only possible once it is measured and monitored accordingly. One first starting point is the area-wide and seamless monitoring of wildlife populations in forests to manage their impact on diverse vegetative structures. Current methods like camera-based traps rely on estimates derived from samples and are unsuitable for large-scale applications. In contrast, camera-based observation using uncrewed aerial vehicles (UAVs) or satellites is suitable for covering large areas. Aerial observation, however, has limits when monitoring in areas with dense vegetation such as forests.

ALFS (Airborne Light-Field Sampling) is a novel approach that supports the removal of occlusion caused by vegetation or other visual obstacles. It is based on light field technology, where a sequence of color and infrared camera images captured from a UAV flight are merged with positional information and a terrain model. The recorded data makes it possible to focus on and visualize objects at specific distances, such as animals on the ground. ALFS has been applied successfully in experiments for detecting persons in search-and-rescue scenarios. In the BAMBI project, an AI-assisted system using ALFS will be created. The system can detect and classify animals in forests and in open terrain to enable comprehensive and accurate monitoring of animal populations.