Remote sensing of eagle nests

Important results from the project

The project "Örnkoll med AI" aimed to train an AI model to recognize features in satellite images that indicate the presence of natural values and at the same time train Ecogain´s staff in the use of AI in their activities. Ecogain has together with Kairos Future developed and trained an AI- model that points out areas with high and low probability for high natural values. The AI model is built to be further developed. There are already map layers in GIS that show areas throughout Sweden with a high probability of high natural values.

Expected long term effects

The result of the project is that employees at Ecogain have gained practical and theoretical knowledge of AI. This is done by practicing developing and learning to use the AI model to identify places with an increased probability of specific natural values. The ultimate goal is not to completely replace the human processes with the AI model, but to integrate the AI-based way of working in the proven working methods and through AI get a better knowledge base.

Approach and implementation

** Denna text är maskinöversatt ** The project plan has been followed with the development and application of the AI model. Experience during development is that high-resolution satellite images gave exponential growth of processor needs. Further experience is that the data warehouses with open biotope data are in many cases old, do not reflect reality, with sometimes changed land use. Land use such as clear-cutting is then not included in the data warehouses. With better quality in input data, the model will be even more accurate