基于多小波分析的图像优化去噪方法研究

Image Optimization De-noising Method Research Based on Wavelet Analysis

针对传统单小波在对称性、正交性、有限支撑等特性上的不足,提出一种基于多小波分析的图像优化去噪方法。对图像进行预滤波处理操作,消除多小波的不恰当离散性。对待处理信号经预滤波处理后产生的四个分量进行多小波变换处理。采用Visu shrink方法与基于stein无风险估计的Sureshrink方法对阈值进行确定。保持总像素量不变,给出多小波分解及重构系统框图。以S=2的多小波为例对多小波变换进行分析。分别完成行滤波和列滤波。对噪声方差进行预测,完成对原始图像多小波系数的方差的估计,对尺度参数和阈值进行计算。仿真实验结果表明,所提方法能够有效去除图像噪声,进一步增强了图像的信噪比。

In view of the traditional unit-wavelet in symmetry, orthogonality, limited support features such as the inadequa-cy that go up, put forward a kind of based on wavelet analysis of image denoising method optimization.Pre filter on image processing operations, eliminating inappropriate discrete multiwavelet.Dealing with the signal after pre filtering treatment on the four components of multiwavelet transform processing.Using Visu shrink method and based on Sure stein, risk-free estimated the shrink method to determine the threshold.The total amount of pixels unchanged, wavelet decomposition and reconstruction system block diagram is given.Based on the wavelet S= 2, for example analysis of multiwavelet transform. Complete line filtering and column filtering respectively.Forecast the noise variance, the completion of the original image wavelet coefficients more the variances of the estimates, the dimension parameters and the threshold value is calculated. The simulation results show that the proposed method can effectively remove the noise of the image, further enhance the SNR of the image.