分数阶小波时频域的图像去噪新方法

An image denoising method based on fractional wavelet transform in time-frequency domain

为了提高图像的质量以及满足后续图像处理的需求,提出了一种基于分数阶小波时频域的图像去噪新方法。该方法通过二维分数阶小波变换将图像映射到分数阶小波时频域内,在时频域内实现图像的去噪处理,最后通过分数阶小波逆变换实现图像的重构。图像去噪实验结果表明:采用该方法去噪后的图像输出峰值信噪比明显提高,在抑制噪声的同时可以有效保持图像细节。

In order to improve the image quality and meet the needs for a higher level of follow-up image pro-cess,a novel image denoising method based on fractional wavelet transform （FWT）is proposed in this paper. The noisy image was transformed into factional wavelet time-frequency domain by a two-dimensional FWT algorithm. Then the transformed image was filtered. Finally, the image was reconstructed by an inverse FWT. The experi-mental results show that the PSNR of output images can be effectively improved. Moreover ,the proposed method can preserve detailed information effectively while reducing the noise.