基于脉冲噪声点检测的中值滤波改进算法

Improved Algorithm of the Median Filter Based on Detection of Impulse Noise Points

在对中值滤波及其他改进算法进行研究分析的基础上,提出了一种基于脉冲噪声点检测的中值滤波改进算法。通过设定阈值标记出可能的噪声点,再根据相邻像素的相关性对可能的噪声点进一步判断,精确确定噪声点,只对噪声点进行中值滤波。实验结果表明,相对于中值滤波及其他改进算法,新算法获得的去噪后的图像具有更好的客观评价指标和主观视觉效果。

Based on the analysis of the median filter and other improved median filters, a new median filter algorithm was pro- posed based on detection of impulse noise points. The possible noise points were marked out by setting threshold value. Then ac- cording to the correlation of adjacent pixels, the possible noise points were further judged and noise points were accurately deter- mined. Only noise point was filtered by noise median filter. Simulation results show that, compared with median filter and other im- proved algorithm, the restored images that are obtained by the proposed new algorithm have better objective quality and subjective vision effect.