バードストライク防止のための鳥種識別機能をもつリアルタイム野鳥検知システム

抄録

Bird strikes at rotating wind turbines have become serious issue due to worldwide exponential spread of wind power. To save wild bird life, especially for rare species registered in red list such as white-tailed eagle, a real time based system to avoid collision in wind farm is strongly desired. We have developed an automatic optical bird sensor applicable to natural background with help of a high-speed image processor coupled with a long-ranged high-resolution camera. In this report, we present how the image processing works for flying bird detection, and also how the bird species are identified to alert turbine-striking probability that is estimated dependent on bird species.