Day 4

Detailed paper information

Back to list

Paper title Combining LiDAR data, 360 imagery and game engines to advance forest monitoring
  1. Raphael Zuercher Wageningen University and Research Speaker
  2. Alexander Klippel Wageningen University & Research
  3. Alvaro Lau Sarmiento Wageningen University & Research
  4. Jiayan Zhao University of Arkansas at Little Rock
  5. Hristina Hristova Swiss Federal Research Institute WSL
Form of presentation Poster
  • C2. Digital Twins
    • C2.01 Towards a Digital Twin of the Earth - advances and challenges ahead
Abstract text Forests are an integral part for the world’s ecosystem, afforestation and deforestation are main drivers for climate change and therefore their monitoring is vital. Forest monitoring involves remotely sensed data, such as Light Detection and Ranging (LiDAR) to capture complex forest structure. Natural environments like forests are complex and add challenges in communication. Conventionally, the forest monitoring data has been analysed in 2D desktop computers, but there is a fundamental shift in this communication due to recent developments in computing and 3D modelling. With the help of game engines and the retrieved forest monitoring data, digital twins can be created.

LiDAR is used to determine exact locations and dimensions of objects. The combination of LiDAR and immersive technologies can be used for stand assessments and measurements and makes it experiential. Further, georeferenced 360-degree immersive imagery and videography complements the abstract LiDAR data with a realistic experience as naturally perceived by the human eye. A workbench provides tools to manipulate the data, including scaling and rotation, but also measurement tools including distance for tree heights, plane for calculating the diameter breast height and volume to approximate the biomass within the immersive virtual reality experience. Satellite imagery with terrain elevation data provides an overview of the research site.

We intend to present the findings of our ongoing research activities in virtual reality forest monitoring and try to answer the questions whether modified meshed LiDAR data measured in virtual reality is as accurate as conventionally measured point clouds and further whether the application helps experts in visualizing and monitoring forests. This is determined with a heuristic evaluation and a usability study. We collected our data in the Eifel national park in west Germany with terrestrial, mobile and drone mounted LiDAR, a Go Pro MAX mounted on a tripod and drones and a microphone. This Beech, Norway Spruce and Oak dominated forest is declared to become a native forest with only minimal human interaction.

The research investigates the benefits and limitations of the single elements of the application, such as the digital terrain models and map, the terrestrial, mobile and airborne LiDAR data, the 360-degree immersive media, the measurement tools, and the forest sounds. An iterative process ensures implementation of feedback from experts. The research further includes exploration of tools, such as using the PlantNet API to use the deep learning model to determine the species of trees with the help of screenshots in the 360-degree imagery within the immersion.