摘要
本文提出了一种新的用于散斑噪声抑制的非线性加权均值多方向广义形态滤波算法。对实际激光雷达图像处理结果表明:本文算法具备了小波阈值算法图像边缘保持好和局部统计滤波算法散斑噪声抑制能力强的优点,在算法复杂度降低的前提下具有了与F.Safa的改进算法(广义并行加权多方向形态滤波算法)相当的效果,即有效地抑制了散斑噪声,又保持了图像的几何结构。
In this paper, a new nonlinear weighted average multidirectional generalized morphological filtering algorithm is proposed to suppressing speckle noise. Processing to the images from CO2 Coherent Laser Radar shows the algorithm has nearly the same edge preserving performance as wavelet soft thresholding algorithm, and nearly the same speckle noise suppressing performance as Lee's filter. This algorithm has nearly the same performance in the edge preserving and speckle noise suppressing as the generalized parallel weighted average multidirectional morphological filtering algorithm that is a modified F.safa's algorithm, and more simple than it. That is say, it can not only effectively suppress speckle noise but also preserve the geometrical features of images.
出处
《信号处理》
CSCD
2003年第2期145-148,共4页
Journal of Signal Processing
基金
国家自然科学基金(批准号:60172038)
全国优秀博士学位论文作者专项资金资助项目(200037)
教育部优秀青年教师资助计划项目(EYTP)
高等学校优秀青年教师教学科研奖励计划(TRAPOYT)
黑龙江省自然科学基金资助
