多尺度全变差图像去噪算法

结合小波和全变差方法，提出了一种多尺度全变差去噪算法. 该算法结合了小波的多尺度结构，并通过对多层小波分解后的高低频施加不同的正则化参数，在频率域上实行多尺度全变差去噪.实验结果证明，所提出的算法相对于全变差去噪方法和小波去噪方法，能够更好地抑制噪声、保持边缘和消除阶梯效应.

A Multiscale Total Variation Algorithm for Image Denoising

A multiscale total variation algorithm is proposed. By taking advantage of the multiscale structure after wavelet transform, the proposed algorithm applies different regularization parameters to low frequency and high frequency blocks. Numerical experiments show that the proposed algorithm eliminates the noise better while preserving the meaningful edges comparing to using Wavelet method or total variation method alone.