A multi-scale morphologic runway detector

In order to enhance the environmental awareness of pilots to some extent, runway detectors play a very important role in Synthetic Vision System. Based on the characters of infrared image, which has bad contrast and heavy noises, an improved multi-scale morphologic edge detector is proposed in this paper. It uses different size of structuring elements to extract the edge points of infrared runway image at first, integrating the superiority of small size detecting fine edges and large size removing more noise. By setting the edge information detected at each different size with different weight just according to their abilities of removing noise, and then combine them together, we can obtain the actual and explicit runway edge. Comparative study reveals its superiority over other traditional edge detectors and morphologic edge detectors. It can detect the runways effectively and precisely which is very beneficial in the aviation areas.