Save birds by Kronomav
How to save birds of prey such as vultures and eagles from wind turbine danger by using a 3D surveillance system?
Birds in danger
Every year big birds of prey smash into wind turbines with a significant number of casualties. In this regard, it is noteworthy that the mortality rate of these birds by collision with wind turbine blades at some points is significantly higher than that caused by other infrastructure such as power lines.
The species often affected by this problem are Griffon vulture (Gyps fulvus), golden eagle (Aquila chrysaetos ), the booted eagle ( Hieraetus pennatus ) or the Great Bustard (Otis tarda).
Therefore, urges the implementation of mechanisms to minimize this negative impact to the maximum and to ensure a sustainable future.
We’ve been asked to build a system to target, monitor and define flight trajectory of these big birds.
The propose of our system is not only to detect birds in flight, but place them in a three dimensional space in order to determine its flight path. Once we’ve determinate the trajectory, if it is in the path of collision with the turbine blades, we could implement deterrence actions.
Kronomav 3D System
We've built a 3D system (two HD camera in visible spectrum) and developed a software tool for detection and image recognition. The system consists of three main parts: 3D cameras, computer and software image analysis.
Cameras or sensors observe an adequate field of vision to include wind turbines and a catchment area in which detect birds and their behaviour.
The system detects in the area to monitor, automatically, the flight of birds and determining the distance to the bird (position x, z ) and the size thereof. With these data, calculated several times per second with the information from the cameras, it determines the flight path through the complementary software and thus emit signals or alarm before a potential collision with a wind turbine.
In addition, the system must record images that can be reviewed later for evaluating, together with the data output of alarms generated, the level of accuracy of the same . Images must likewise provide sufficient information to assess / compare the results of the monitoring of birds that traditional will take place during the first years of operation of the system.
The system incorporates a software analysis of birds present in the area to be analysed, by tracking the flight path to determine a possible collision on wind turbines in the wind farm present.
* Position and calibrate a set of two cameras to obtain a three-dimensional measurement system with high resolution.
* Analyse the images from these two high-resolution cameras to detect and determine the trajectory of the birds of prey .
* Identify and anticipate the direction of flight of birds in case of passage through the swept area of the turbine blades, causing a possible collision alarm .
* The alarms generated, along with the paths of previous movements to the alarm are captured on videos and images.
* Every time a bird enters the alarm field of view, generation camera images with higher resolution are stored to proceed to the subsequent identification of the species.
We have faced some big challenges:
* Target small objects (birds) in a moving and constantly changing landscape.
* Define the optimal placement of the system to determine the influence area (danger area) where animals or objects are detected.
* Eliminate false positive targets, birds under 1 meter long are discriminate.
* Once a bird of prey is target, define the trajectory
We have met all the challenges the project proposed, now we’re able to save the birds by sending dissuasive signals or changing the turbine blades position.