Application of the directional wavelet transform in edge detection of airfield runway

The airfield runway usually takes on directional orderliness in the aerial images and remote sensing images, but the traditional edge detection algorithms do not reflect this feature. The directional wavelet transform, which drawed in a directional angle variable thetas, carried out ldquonon-maximum suppressionrdquo and ldquohysteresis thresholdingrdquo on the pixel that through the transform of directional wavelet. It could extract the edges in certain orientation, simultaneously restrain interferential objects in other directions. This approach was employed to detect the edges of the airport runway in the airscape and remote sensing images. Experimental results indicate that this method is insensitive to noise, less computability, and the edges detected have good continuity and localization. Therefore, the method is appropriate for processing such kinds of images.