【摘要】 图像信号滤波是数字图像处理领域的重要技术之一。传统的滤波方法存在去除噪声却引起图像边缘模糊的问题,利用GHM多小波变换分析图像信号和噪声的性态,结合贝叶斯估计方法进行非线性阈值和重构,实现了图像信噪分离的目的,提高了信噪比。仿真结果表明,该方法能够有效地抑制噪声,且较好地保留了图像细节。更多还原

【Abstract】 The filtering of the image signal is one of the important technologies in the filed of digital image processing. The traditional filtering method lies in one problem of being brought image skirt faintness when it cleans off noise.The characteristics and performance of image signal and noise are analyzed based on GHM multiwavelets, reconstructs the image with the non\|linear threshold method combined with Bayes estimation, separates the image signal and noise. The simulation result shows that the method can restrain the noise efficiently and keep the high frequency detail part very well.更多还原