Community driven approach to 3D hyperspectral AI for universal biodiversity classification
| Reference number | |
| Coordinator | Linköpings universitet - Linköpings Universitet Inst f systemteknik ISY |
| Funding from Vinnova | SEK 1 995 000 |
| Project duration | October 2023 - September 2026 |
| Status | Ongoing |
| Venture | Advanced digitalization - Enabling technologies |
| Call | Advanced and innovative digitalization 2023 - call two |
Purpose and goal
The EU´s biodiversity strategy for 2030 represents an ambitious long-term plan aimed at protecting nature and reversing ecosystem degradation. To facilitate the realization of this plan, this project aims to develop a 3D Hyperspectral Imaging Artificial Intelligence (3DHSIAI) system for universal biodiversity classification. By combining advanced imaging techniques, AI algorithms, and autonomous drones, the project will create a powerful tool for accurately and efficiently identifying and classifying various species and habitats across different ecosystems.
Expected effects and result
The expected effects and results include: 1) A standardized hardware solution for the generation of 3D hyperspectral data. 2) Novel 3D AI algorithms that can leverage such a hardware solution for biodiversity classification. 3) A large-scale open-source GitHub community for 3D biodiversity classification.
Planned approach and implementation
The project will be structured into six distinct work packages to achieve hardware development, data collection, and algorithm design. It will be completed within a 3-year timeframe, including the development, optimization, and validation phases.