基于人类视觉的B样条滤波算法  

B-spline filter algorithm based on HVS

作　　者：潘梅森[1] 颜君彪[1] 肖政宏[1] 

机构地区：[1]湖南文理学院计算机科学与技术系,湖南常德415000

出　　处：《计算机工程与设计》2007年第6期1364-1367,共4页Computer Engineering and Design

基　　金：湖南省教育厅资助科研基金项目(05C720)

摘　　要：提出了一种基于人类视觉系统的B样条滤波算法,首先根据人类视觉特性确定噪声点,然后根据3×3窗口内噪声点个数自适应调整滤波窗口大小,最后采用B样条函数和中值滤波方法对噪声点进行逐点滤除。该方法既能有效地去除噪声,又能较好地保持图像细节部分。最后给出模拟实验和分析,结果表明该方法是有效的,较标准均值滤波和中值滤波具有更好的滤波性能。A B-spline filter algorithm based on human vision system is proposed. Firstly the noise pixel is detected according to a humanity visual characteristic, then the filter window size is adaptively adjusted according to the window size of 3×3, finally the b_ spline function and median filter are used to estimate noise pixels. This method both effectively remove the noise, and maintain the image detail well. Finally, the experiments and the result analysis are obtained. The result shows that this algorithm is effective and has better filtering performance than the standard mean filter and median filter.

关 键 词：椒盐噪声 B样条 人类视觉系统 峰值信噪比 中值滤波 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]