Drones in Forest Inspection

In recent years, forestry applications have seen an increased use of drones in forest inspection. Having a highly dynamic nature, forests need regular and timely data acquisition, especially when it comes to bark beetle infestation. As it can spread very quickly and result in great losses, use of drones in forest health monitoring is becoming essential. It enables fast and effective response already during early stages of bark beetle occurrence.

Whereas conventional remote sensing techniques have often proven insufficient in such cases, UAVs can operate with flexibility and yield data in high spatial resolution. It is also possible to study the dynamics of bark beetle attacks on the basis of maps generated from photographic surveys combined with airborne images and satellite data to improve early bark beetle detection. Last but not least, some national parks have used UAV-based data for regular forest monitoring and inventory, e.g. to map dead wood on the forest floor.

Fly4Future participated in the KOLDRO project that developed solutions for forest inspection and elimination of European spruce bark beetle infestation.

This solution involves two autonomously collaborating drones – a detecting drone and a marking drone. The detecting drone equipped with a multispectral camera detects the infested trees while flying above the forest. When an infested tree is discovered, the information is sent to the marking drone.

The marking drone then enters the forest. It scans the tree’s bark and phloem for small holes in the bark beetle using a high-resolution RGB camera and an algorithm based on deep learning.

After the tree trunk segmentation, hole detection, and final classification the 3D lidar is used for the accurate image capture and measurement of the distance to the tree. If the bark beetle is confirmed, the drone uses its spraying device to mark the tree with paint.

Thanks to the KOLDRO project and involvement in other innovative applications, Fly4Future gained significant expertise in developing autonomous or semi-autonomous systems for various purposes. These systems are capable of efficiently surveying vast areas of forests, effortlessly navigating through dense vegetation, and collecting precise and valuable data, including the identification of the significant presence of dead wood on the forest floor. This enables a comprehensive analysis of ecosystem health, aiding in biodiversity assessments and habitat mapping. By harnessing the power of automation in forest monitoring, we contribute to the development of more effective and sustainable resource management strategies and environmental conservation efforts.