一种稳健快速的椒盐噪声抑制算法

针对图像中椒盐噪声的特点,提出了一种细节保持的椒盐噪声的滤波算法.算法分两步进行:第一步先对噪声图像进行自适应中值滤波,而后使用逻辑判断的方法对可能噪声点进行标识;第二步采用了一种新的逐次修复的方法,利用当前噪声点周围的非噪声点或已修复的噪声点的信息对当前的噪声点进行修复.我们对密度在10%~95%的各种噪声强度的情况进行了仿真,结果显示本文提出的算法能在有效地抑制噪声的同时完好地保持图像的细节信息,尤其在高噪声密度条件下显示了良好的噪声抑制效果和细节保持能力,显示了算法的稳健性.文中还给出了一种相当有效的优化策略,极大地降低了算法的运算复杂度,使算法适合于实时实现,有实际应用价值.

A Reliable and Fast Salt-and-pepper Noise Removing Algorithm with Detail-Preserving

This paper proposes a two-phase scheme for removing salt-and-pepper impulse noise. In the first phase, an adaptive median filter based algorithm is used to identify pixels which are likely to be contaminated by noise. In the second phase, the image is restored by a stepwise method which repairs each noise pixel using its eight closet neighbors. Simulation is be done for all various noise density from 10% to 95%, and the results show that the proposed algorithm can restrain salt-and-pepper noise with detail-preserving, especially for high noise level.Also this paper proposed some efficient optimizing strategy which reduces the computation complexity. This make the proposed algorithm fit for real-time application.