摘要
根据航海雷达数据的时空特点,提出一种图像噪声检测与滤除方法。该方法首先根据噪声在各次扫描间的不相关性对其进行检测,然后只针对检测出的噪声数据进行滤除。通过与传统的中值滤波算法和均值算法进行比较,同时使用信噪比指标对处理结果进行分析,结果表明,本算法既保持了传统中值滤波方法简单、快速的优点,又很好地保持了图像的灰度信息,具有较好的降噪效果。
A novel algorithm for noise detection and filtering is proposed in this paper based on space - time characteristics of nautical radar data. The noise can be detected and filtrated by the improved median filtering algorithm according to the uncorrelated property among various scans. The algorithm for traditional median filtering and that for mean filtering are analyzed in comparison with the algorithm proposed in this paper, and the evaluation parameter of SNR is utilized to estimate the results. It is shown that the algorithm proposed in this paper preserves not only such advantages of the traditional median filtering as simple calculation and high speed computability but also the gray information of the original image. In addition, a good noise reduction effect can be achieved.
出处
《国土资源遥感》
CSCD
2008年第2期14-17,共4页
Remote Sensing for Land & Resources
关键词
航海雷达
中值滤波
同频干扰
信噪比
Nautical radar
Median filter
Shared- frequency- jamming
Signal noise ratio（SNR）
