We put forward a process of automatic airport extraction based on the characteristics of high resolution remote sensing images. First, through image enhancement algorithm, the contrast of target and background is enhanced. Second, we can extract the possible airport through the algorithms of Ostu segmentation, mathematical morphology corrosion and region-labeling. Finally, combined with the geometric structure of the runway, the airfield runway can be extracted through the algorithms of edge detection, progressive probability Hough transform and line connection. Then the possible airport can be verified by the extracted airfield runway. In the process, we proposed an improved fuzzy enhancement algorithm for image enhancement. This algorithm has good effect on the image enhancement and has strong robustness. The results of the experiment indicate that the process of automatic airport extraction is robust and has the advantages of high speed and degree of automation.