节点文献
遥感图像中机场识别与毁伤分析研究
【作者】 范瑞彬;
【导师】 夏德深;
【作者基本信息】 南京理工大学 , 计算机软件与理论, 2004, 硕士
【摘要】 机场识别技术是计算机模式识别领域逐渐兴起的研究课题,在军事上有着重要的应用前景。机场识别问题也是模式识别领域的一个相当困难的问题,要使这一技术成为完全成熟的技术还有许多工作需要去做。本文结合已有的一些方法,对于机场识别的若干问题做了一些研究和改进,取得了一些有意义的结果。 本文工作包括: (1)从机场检测着手,针对实际情况,采用了一种新的基于直线特征的机场检测方法。首先采用Canny算子提取图像中的边缘,再采用链码跟踪算法提取出直线边缘,然后根据先验知识建立判断规则,将提取出的直线边缘进行编组判定,可以准确的确定机场存在的区域。这种方法减少对跑道的灰度特征的依赖,更多利用跑道的形状特征,形状特征受光照、噪声、不同时相等影响要比灰度特征小的多。 (2)在已检测出机场的基础之上,介绍了一种新的有针对性的跑道识别方案。先对机场区域图像进行灰度形态学增强,再用基于矩的阀值选取法进行图像分割,然后再用像素标记法提取出机场连通域,最后采用Hough变换识别出机场中的跑道。 (3)基于军事应用的要求,不仅要准确检测和识别机场目标,还要能够分析机场在毁伤打击后的变化情况。所以,这里根据毁伤打击前后的图像对机场目标的毁伤变化情况进行相关性匹配从而检测出毁伤区域,同时还对基于毁伤打击后图像的弹坑检测作了初步的研究。
【Abstract】 The technology of airport recognition is an active subject in the area of pattern recognition. There are broad applications in the military affairs. For the particularity, the airport recognition is also the very difficult problem. There is still much work to do. In this paper, we make some researchs and progress on the airport recognition based on some current algorithms, and achieve some results.The work including:(1) This paper introduces a new airport detection approach based on many regular straight line features in airports. Firstly , we detect the edges of the remote sensing image, and then, extract straight line chain codes. Finally, we group the straight line chain codes, and confirm the airport area by the rules we made. The experimental results demonstrate that the method of airport detection is valid. This method is more lied on the shape features of airports rather than the gray value features, because the shape features are more robust than the gray value features.(2) This paper presents a new runway recognition approach based on the above step. Firstly, we used morphologic enhanced the airport images. Secondly, we used the based-moment threshold selection to segment the image., and then , extract the connected component of the airport with a pixel labeling algorithm .Finally, we recognize the runway with the Hough Transform.(3) This paper introduces two methods to analyses the airport damage post strikes. One is change detection algorithm based on image matching, the other is craters recognition. Then we point out the merits and the limitations of the two methods.
【Key words】 Airport Recognition; Airport Detection; Runway Recognition; Damage Analyses;
- 【网络出版投稿人】 南京理工大学 【网络出版年期】2004年 04期
- 【分类号】TP751
- 【被引频次】14
- 【下载频次】605