一种基于新型符号函数的小波阈值图像去噪算法

在现有阈值去噪算法的基础上提出了一种基于新型符号函数的小波阈值图像去噪算法,该算法提出的新阈值函数具有连续可导、小波系数偏差小、阈值自适应性强等优势.不仅保留了分解后的低频小波系数,还有效滤除了高频系数中的噪声系数,使得重构后的图像更接近原始图像.对高斯白噪声的Bridge图像、Lena图像及含“斑点噪声”的B超Fetus图像进行仿真,实验的结果表明,无论是新阈值函数的视觉效果,还是定量指标PSNR和MSE,均优于现有的阈值图像去噪算法.其边缘及细节信息能得到较好的保护,无明显振荡,图像更平滑、均匀,且在复杂噪声背景下,该方法具有较好的顽健性.

A wavelet threshold image denoising algorithm based on a new kind of sign function

Based on the existing threshold denoising algorithm,a threshold denoising algorithm based on the new symbolic function was proposed.The new threshold function has advantages of continuous guidance,small deviation of wavelet coefficient,strong threshold adaptability and so on.It not only preserved the low-frequency wavelet coefficients,but also filtered the noise coefficients in the high-frequency coefficients effectively,so that the reconstructed image was closer to the original image.The simulation results of Bridge image,Lena image and B-mode Fetus image with Gaussian white noise show that the visual effect of both the new threshold function and the quantitative indicators PSNR and MSE are better than the existing threshold image denoising algorithm.The edge and detail information can be better protected,have no obvious oscillation,the image is smoother and even,and the method has good stubbornness under the background of complex noise.