基于异性扩散-中值滤波的超声医学图像去噪方法

ULTRASONIC MEDICAL IMAGE DENOISING METHOD BASED ON ANISOTROPIC DIFFUSION-MEDIAN FILTERING

针对超声图像存在一种特殊的斑点噪声,使图像边界与细节变得模糊而严重影响图像质量的问题,提出了一种新的去除医学图像斑点噪声的方法,它利用中值滤波和各向异性扩散相结合,不仅可以有效地去除噪声而且很好地保持了边缘、局部细节信息。此外,该方法在扩散过程中,梯度阈值选取的不同对图像结果影响很小,这极大地提高了该算法的健壮性。实验中,通过和各向异性扩散、中值滤波等方法的比较,表明该方法具有良好的去噪效果。

A kind of noise called speckle in ultrasonic images degraded severely the sharpness of image, especially edges and details. A new approach against this problem which reduces speckle noise and enhances structures in speckle-corrupted medical images was proposed. It utilized a median fihering-anisotropic diffusion compound scheme. This hybrid approach achieved much better noise suppression with minimum edge blurring while kept the edge and detail information of ultrasound images well. In addition, the performance of the result was less sensitive to the image gradient threshold selection during diffusion and that reflected the enhanced robustness of the algorithm. The denoising effect of the proposed technique is clear when it is compared with the anisotropic diffusion algorithms or the median filtering algorithm only in medical image.