Airport Birds Optical Image Tracking Identification Technology Research

The prevention of birdstrike was a worldwide problem of aviation.Recently with the develepment of economy, the birdstrike events increased with an increase in flights and airlines. To protect the aircrafts' safety, the way to control birdstrike effectively was becoming more and more important. Compared with the traditional anti-birds system by timing or artificially operation, the technology of video tracking recognition of birds could identify birds intelligently and implement the anti-birds according to the results of the image recognition. It could also reduce the birds' habit of some kind of anti-bird's signal observably and improve the effectivity of anti-bird. Meanwhile it could take notes on the occasion of the birds in the airport and set up homologous database which was beneficial to the improvement of the work for anti-birds.The major contents in this thesis included investigation of birds tracking recognition and drive technology, the image recognition program by OpenCV and image processing operations using the functions in OpenCV library. The details are described as follows:1) Pretreatment of the video image ahead of time, including graying,noise reduction and morphological operations.2) Establishment of the adaptive background model and the recognition of birds.3) Based on the randomness of birds' flying, a comparative model of image was put forward to identify birds easily.4) In the end, the program was tested with the video of the birds over the airport and analyzed totally.It showed that a good robustness between the experiment and the program was obtained, which was able to identify the birds on time.