二维分数阶小波变换滤除混合图像噪声研究

图像形成与传输过程中,常受到复杂混合噪声的干扰.本文结合二维分数阶小波变换与中值滤波,提出一种新的混合未知图像噪声滤除方法.该方法先通过噪声检测将脉冲噪声标识出来,并利用中值滤波方法滤除,然后计算剩余噪声,进一步得到二维分数阶小波变换的最优阶次,在二维分数阶小波时频域,将剩余的高斯白噪声用阈值去噪方法滤除.经实验证明,该方法在有效去除混合噪声时具有优势.

Research on Mixed Noise Filtering in Image Based on the 2-D Fractional Wavelet Transform

In the forming and transporting process,images are often disturbed by complicated mixed noises.Combined the 2-D fractional wavelet transform with median filtering,a new image filtering method for filtering mixed and unknown noise was proposed.First,impulse noise was detected and marked up by the impulse noise detection,and filtered by median filtering in this method.Then the residual noise was calculated,further the optimal fractional order of the 2-D fractional wavelet transform can be got,and the residual Gaussian noise can be removed by using the threshold shrinking method in 2-D fractional wavelet time-frequency domain.The experiments show that the proposed method can remove the mixed noise effectively.