航空遥感图像中道路检测方法研究与仿真

Airborne Remote Sensing Image Road Detection Method and Simulation

研究航空遥感图像中道路准确检测的问题。航空遥感图像中,道路是一类复杂目标.道路目标不同于一般刚体,它难以用固定的参数描述,只能用一些抽象的语句来描述.传统的检测算法多是对一些固定的识别目标完成描述,用道路图像非固定特征参数来完成大量复杂道路目标的检测是非常困难的。提出了一种基于主元分析算法的航空遥感图像中道路检测方法。利用LBP方法进行特征提取,从而获取航空遥感图像中的道路特征,为航空遥感图像道路检测提供准确的数据基础。利用主元分析方法对所有的像素进行重新排列,通过二进制方法描述获取的对比结果,提取动态特征。实验结果表明,算法极大地提高了检测的准确性,使获取的结果与实际道路情况相符。

Road accurately detect problems in the research aviation remote sensing image.Aviation remote sensing image,the road is a kind of complex target.Target is different from general rigid road,it is difficult to use a fixed parameter description,can only use some abstract statements to describe.The traditional detection algorithm is more complete description on some fixed to identify the target,with the road image without a fixed characteristic parameters to accomplish a large number of complex targets detection is a very difficult road.Put forward a kind of aviation remote sensing image based on principal component analysis algorithm of road detection method.Using LBP feature extraction method,thus obtaining way in aviation remote sensing image characteristics,provide accurate data for road detection aviation remote sensing image.Way to make use of principal component analysis method,to identify,so as to complete aviation road detection in remote sensing image.The experimental results show that the algorithm presented in this paper for road detection in aviation remote sensing image,greatly improve the accuracy of the detection,to obtain the results more consistent with the actual road conditions.