摘要
针对现有脉冲噪声滤波算法在处理大斜率区域和断线型激光扫描数据中存在的多判和漏判问题,提出了一种有效检测脉冲噪声的方法.该方法依据局部脉冲噪声的分布,利用自适应弦偏差法进行噪声点检测;对噪声点自适应地选择滤波窗口中的非脉冲噪声数据进行中值滤波,以实现噪声滤除.实例结果表明:采用文中方法使得噪声检测更为准确,在滤除噪声的同时有效地保持了数据的细节.
According to the distribution of local impulse noise, an adaptive chord deviation method is used to avoid redundant detection or missing of impulse noise at the quick rising, descending or broken line area of data set. It uses median filter to filtering out impulse noise by adaptively tackling the non-impulse-noise data in filter window. Test examples show that the proposed method outperforms some known methods in noise reduction and detail preservation.
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2006年第10期1531-1534,共4页
Journal of Computer-Aided Design & Computer Graphics
基金
国家自然科学基金(50175063)
