Digital aerial surveys of seabirds and marine mammals in the Netherlands
18 December 2024
The roadmap ‘Automatic analysis of digital aerial surveys’ presents a multi-phase strategy to automate marine animal surveys using digital aerial imagery and AI-based image recognition. The aim of this initiative is to enhance the detection and identification of seabirds and marine mammals, focusing on high-definition imagery and deep learning to increase efficiency, accuracy, and cost-effectiveness:
Traditional seabird and marine mammal surveys are labour-intensive and cannot be done above offshore wind farms. Automated digital aerial surveys can increase efficiency by processing large volumes of high-definition images quickly, reducing reliance on human observers and enabling more consistent data collection.
By switching to automated image analysis, the process becomes more transparent. Images can be re-examined and verified, unlike human visual surveys. Automation will also significantly lower the costs associated with marine surveys. By reducing manual effort and enabling continuous, large-scale monitoring, automated systems make it feasible to expand surveys to cover more extensive geographic areas and longer timeframes. This scalability is crucial for tracking marine populations in an era of increasing environmental pressures.
Open-source, AI-driven platform for automated detection and identification
The roadmap outlines a possible approach to develop an open-source, AI-driven platform that can automatically detect and identify marine species in aerial images, supporting tasks like population monitoring and environmental impact assessments. It describes three phases in the development of such a tool: short-term (proof of concept), mid-term (platform expansion), and long-term (global open-source deployment).
Automated digital aerial surveys could be a transformative tool for marine biodiversity monitoring, offering significant improvements in environmental data collection and analysis. By integrating AI with digital imaging, this roadmap paves the way for smarter marine resource management.
Read the full report here
This roadmap has been developed by Wageningen Marine Research, Observation International, Waardenburg Ecology, Rijkswaterstaat CIV Datalab, and Vision+Robotics on behalf of Rijkswaterstaat.
Read more about our research on using Artificial Intelligence to recognise seabirds and marine mammals in and around offshore wind farms.
