椒盐噪声下图像原始直方图估计的收敛性实验

Estimation of original histogram for images under salt and pepper noise——An experiment on convergence

获得噪声干扰图像的原始直方图有助于判断像素的原始灰度值.分析了脉冲噪声下图像的原始直方图估计出现局部偏差的原因,并以噪声率给定而噪声散布逐步趋于均匀的图像序列作仿真,实验结果表明随着脉冲噪声趋于均匀成为椒盐噪声,图像的估计直方图将重合于原始直方图.

It would be helpful to estimate the original histogram of an image polluted by noise in recovering its grey values. The case of local biases occuring in the estimated version of the original histogram is analyzed. A simulation is then made using an image sequence of a pulse noise scattering at a given rate and with a progressive evenness. Experimental results show that when the pulse noise scatters sufficiently even to be a salt-and-pepper noise, estimated versions will converge to the original version of the image histogram.