摘要
斑点噪声是合成孔径雷达(synthetic aperture radar,SAR)相干成像所固有的,且不可避免的.为了尽可能地抑制斑点噪声对SAR图像变化检测的影响,利用分形维数方法对SAR图像的变化检测进行了研究,并提出了一种基于SAR图像分形维数的变化检测算法,该算法对SAR图像斑点噪声不敏感.并对计算分形维的滑动窗大小的选择进行了研究.用实测SAR图像进行了验证实验,结果表明分形维数可以用于SAR图像变化检测,它是一种新的SAR图像变化检测的方法与途径;同时,滑动窗大小的选择对SAR图像检测结果有影响.
Speckle noise is inherent and unavoidable for synthetic aperture radar （SAR） images, because its principle is coherent imaging. In order to reduce affection of speckle noise to SAR image change detection as much as possible, fractal dimension method is used to study SAR image change detection. A new change detection algorithm based on SAR image fractal dimension, which is insensitive to speckle noise is proposed. The choice of sliding window size for counting fractal dimension is also studed. Some real SAR images are implemented to verify the method. Experimental results indicate that fractal dimension can be used to perform SAR image change detection and that the new SAR image change detection method--fractal dimension is ideal. At the same time, the choice of sliding window size brings some affection for SAR image change detection.
出处
《西安工业大学学报》
CAS
2008年第5期466-470,共5页
Journal of Xi’an Technological University
关键词
SAR图像
分形维数
变化检测
盒维数
SAR image
fractal dimension
change detection
box dimension
