摘要
针对高密度椒盐噪声污染图像的去噪问题,提出了一种有效的非线性滤波算法。在噪声检测中利用域值划分法,先将噪声图像像素点分为信号点和准噪声点,建立噪声矩阵,然后再利用图像边缘特性及局部统计信息,进一步明确噪声点。对于噪声点,采用以该点为中心的多窗口像素点中值及该点像素值的中值进行替换。实验结果表明,该算法对较高密度椒盐噪声污染图像的去噪能力较强,而且有效地保持了图像细节。
An efficient nonlinear filtering algorithm was presented to remove high-density salt and pepper noise.The threshold classification method is used to divide noise image pixels into quasi-noise point and signal point,meanwhile,to establish the noise matrix,and then to apply the image edge features and local statistical information to further clarify the noise points.For the noise point,its pixels value will be replaced by the median value of pixel set which include the center point and every median of sub-window around center point.Simulation experiment results show that the algorithm has better denoising ability against high-density salt and pepper noise pollution image,and maintains image detail effectively.
出处
《渤海大学学报(自然科学版)》
CAS
2011年第4期362-366,共5页
Journal of Bohai University:Natural Science Edition
基金
国家自然科学基金项目(No:11171042)
关键词
高密度椒盐噪声
非线性滤波算法
噪声检测
多级中值滤波
high-density salt and pepper noise
nonlinear filtering
noise detection
multilevel median filter