遥感图像中的军用机场识别算法研究

提出了一种针对遥感图像的军用机场快速识别方法.首先,对图像进行适当的增强处理,并在此基础上使用Hough变换提取出机场的线状几何特征.然后,利用跑道边缘的平行直线特征和跑道边缘与可能机场区域的重叠关系识别出军用机场.本文用真实的遥感图像测试了这种新算法,实验结果表明该方法能够提供准确的识别结果.

Recognition of Military Airport in Remote Sensing Images

Object recognition is one of the most important and challenging issues in image understanding.In this paper,a novel method is proposed to efficiently recognize military airport in remote sensing images.First of all,lines are extracted from the preprocessed image by adopting the Hough Transform.Then,the recognition is achieved based on geometry characteristics of the airport,such as the paralleled edges of runway,the superposing location between the edges of runway and the airport region.The proposed approach has been tested on real remote sensing images,and the satisfied experimental results demonstrated that the new method can be successfully applied to the recognition of the airport.