Community driven approach to 3D hyperspectral AI for universal biodiversity classification

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 results and effects: 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.

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.