Recognition of airport based on runway-detection

While detecting airport in remote sensing images,the runway often is the first object to detect.In the paper,the binary edge image of airport is first extracted,and then an improved method of Hough transform is used to detect a pair of parallel lines corresponding to runway.The central line of runway is identified based on the pair of parallel lines.Making the pixels of the central line as seed points,the structure of airport can be identified by region-growing.The result of experiment has proved that this method performs well.