一种新的超声图像斑点噪声抑制方法

斑点噪声是超声图像中固有的噪声。现有的用于斑点噪声抑制的自适应滤波方法，小波软阈值方法及小波域内细节抛弃法在去除噪声的同时，不同程度地丢失了一些图像细节。针对这一问题。本文提出了一种新的结合自适应中值滤波和小波软阈值处理的超声图像斑点噪声抑制方法。对计算机仿真图像及超声图像进行处理的结果表明，本文提出的新方法在有效去除斑点噪声的同时，很好地保留了图像的细节，优于上述的其他方法。

A NOVEL ULTRASOUND SPECKLE SUPPRESSION METHOD

Speckle is an inherent noise in ultrasound images. The existing methods such as the adaptive weighted median filter, the normal multi-scale thresholding method and the simple discarded method in wavelet domain failed to trade off between the speckle suppression and the details preservation. To solve this problem, we presented a novel ultrasound speckle suppression method that took the advantage of combining the nonlinear thresholding technique in wavelet analysis with the preprocessing of adaptive filter. A computer simulated image, an in-vitro B-Scan image and an echocardiogram were processed with this novel method. The results showed that the new method was better than other three methods. It could not only effectively reduce the speckle noise but also preserve the resolvable details.