混合噪声环境下一种图像降噪的新算法  

A Novel Algorithm for Image De-noise in Mixed Noise Environment

作　　者：张昊慧[1] 

机构地区：[1]淮阴师范学院物理与电子电气工程学院

出　　处：《电子技术（上海）》2012年第12期27-29,共3页Electronic Technology

摘　　要：针对某些图像中背景复杂,目标对比度和信噪比低等特点,提出了一种基于邻域阈值分类的小波域去噪和改进的中值滤波相结合的方法对图像的混合噪声进行降噪。仿真实验结果表明,新算法能够有效去除高斯白噪声和脉冲噪声。新算法去噪后图像的峰值信噪比和均方误差都比单一算法得到了改善,从而证明了该方法的有效性。Aiming at the characteristics such as complex background, low contrast, and high background noise in some images, a method is proposed to de-noise the mixed noise in the images, which combines wavelet transform based on Neighbouring threshold classification and improved median filter. This method can effectively eliminate the fixed noise of Gaussian White Noise and impulse noise. The simulation experiment shows that the peak SNR and mean square error after de-noising of the images with proposed algorithm is improved over single algorithm, and the validity of this algorithm is proved.

关 键 词：混合噪声 改进的中值滤波 邻域阈值 小波变换 

分 类 号：TN911.73[电子电信—通信与信息系统]