基于线特征的机场跑道定位算法  

Improved algorithm of airfield runway localization based on linear feature

作　　者：杨超[1] 朱敏[1] 王昭莲[2] 

机构地区：[1]四川大学计算机学院,成都610065 [2]重庆通信学院信号与信息处理实验室,重庆400035

出　　处：《计算机应用》2011年第A01期52-54,共3页journal of Computer Applications

基　　金：重庆自然科学基金资助项目(CSTC;2005BB2207)

摘　　要：提出了一种基于线特征的机场跑道识别定位算法。首先,将方向小波变换用于目标的边缘检测,得到了定位准确、边缘连续、背景较为干净的机场边缘图;然后,利用标定了梯度方向的边缘点作为Hough变换的基本元素,将Hough变换的全局"投票"过程变为目标方向上的局部投票,并且将直线段的检测和定位过程融为一体,从而得到跑道的描述信息。实验结果验证了该算法的有效性和稳健性。An algorithm of airfield runway recognition and localization based on linear feature was presented in this paper.Firstly,directional wavelet transform was adopted to obtain edge map with precise localization,continuous edge and clean background.Then,the edge dots marked by the gradient orientation were regarded as the basic elements of the Hough transform.The full-scale ＂voting＂ of Hough transform was replaced by partial ＂voting＂ of target direction.Besides,the straight line detection and localization process were combined together,so that the runway description information could be achieved.Experimental results indicate that the proposed algorithm is effective and robust.

关 键 词：线状特征 方向小波变换 HOUGH变换 直线段定位 

分 类 号：TP391.413[自动化与计算机技术—计算机应用技术]