高分辨率航空影像中高压电力线的自动提取  

Automatic Extraction of Power Lines from Aerial Images

作　　者：李朝阳[1] 阎广建[2] 肖志强[2] 李小文[2] 郭军[1] 王锦地[2] 

机构地区：[1]北京邮电大学信息工程学院模式识别实验室,北京100876 [2]遥感科学国家重点实验室,北京师范大学遥感与地理信息系统研究中心,地理学与遥感科学学院,北京100875

出　　处：《中国图象图形学报》2007年第6期1041-1047,共7页Journal of Image and Graphics

基　　金：国家高技术发展研究"863"项目(2003AA134040)

摘　　要：随着航空数字摄影测量技术和数字相机分辨率的提高,数字摄影测量应用于电力线监测已成为可能。然而,目前却很少有公开发表的关于从航空影像中自动提取电力线的文献见诸报道。为此,提出了一种在复杂自然背景条件下,从航空影像中自动提取550kV高压电力线的算法,并设计了抗强噪声的线特征提取算子,即首先用Ratio算子从遥感影像上提取电力线像素点,然后采用分段Radon变换提取并连接各分段电力线,再利用类似卡尔曼滤波技术跟踪连接电力线的断裂部分,实验表明,该方法能有效地自动提取复杂自然背景下的高压电力线。With the development of aerial photogrammetric technology and the digital camera＇ s spatial resolution. It is possible to use photogrammetric technology in the inspection of power lines. However, there is little published method for automatic extraction of power lines from aerial images. In this paper, an algorithm of automatic extraction of 550kV power lines from complex background in aerial images is presented. Linear feature operators are designed to resist the strong noise and the potential power line pixels are acquired by Ratio operator. Part Radon transform is used to acquire and link the segments. The gaps of the power line segments are filled through similar Kalman filter tracing method. The proposed method is validated by the experimental images.

关 键 词：高压电力线 高分辨率航空影像 自动提取 RADON变换 卡尔曼滤波 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]