Abstract: As one of the problems in the pattern recognition domain, the airport recognition has a significant application prospect in the military affairs. It contains two aspects：identification of airport and localization of runway. A new method to solve the problem of airport recognition is presented. The method of threshold value division is applied to the image, and the pixels marking method is used to extract the maximum cross-connection domain as possible airport area（ROI）, then Canny operator is used to extract the edge of the image. By using an improved Hough transform method, straight line can be precisely extracted from the edge image and locate the airport runway.

Key words: edge detection, Hough transform, airport recognition

摘要： 机场识别作为模式识别领域的问题之一，在军事上有着重要的应用前景。它包括判断机场是否存在和机场定位两个方面。结合已有的方法，提出了一种新的机场识别的解决方案。该方案先使用阈值分割方法进行图像分割，然后利用像素标记法提取出最大连通区域，作为疑是机场区域（ROI）；用Canny算子进行图像的边缘提取，提出了一种改进的Hough变换，能够从边缘图像中准确地提取出直线段，并最终实现机场跑道的定位。

关键词: 边缘检测, Hough变换, 机场识别