flowerpower

Project FlowerPower addresses the critical issue of biodiversity loss in the Netherlands, where approximately 85% of natural flora has been eradicated due to agricultural practices and urban expansion. Recognizing that understanding and measuring biodiversity is essential for conservation efforts, the project aims to leverage artificial intelligence (AI) and drone technology to enhance biodiversity monitoring.

As part of a larger consortium initiative, "Data-driven Biodiversity Management as a Service for Cities," supported by the Lectorate of AI & Big Data and SPARC, Fontys ICT plays a pivotal role in developing tools that facilitate data-driven insights into biodiversity. The project seeks to automate observations of biodiversity, moving away from traditional manual methods that are often resource-intensive and impractical for comprehensive monitoring. While crowdsourcing observations presents an alternative, it is prone to biases related to human preferences for specific times, locations, and species.

To overcome these challenges, FlowerPower explores the integration of drone photography with advanced object detection models. This approach not only enables the capture of data from hard-to-reach areas but also enhances the accuracy and efficiency of biodiversity assessments. Ultimately, the project aims to contribute to a digital transformation in biodiversity monitoring, making it more effective and accessible for conservation efforts.

For more information:

my contribution

For the FlowerPower project I trained an AI model on the Eindhoven Wildflower Dataset to enable automated wildflower counting.

Photo of Bas and Gerard at Naturalis
Bas and Gerard presenting FlowerPower at Naturalis Biodiversity Center
Photo of Bas demonstrating AI model
Bas demonstrating an early version of the AI model

media

Article by the Data Archiving and Networked Service

Article by BRON