摘要:
Because of the existence of speckle noises, classical edge-enhanced operators will result in strengthening the speckle noises of medical ultrasound images and degrading the contrast resolution meanwhile. In view of this, we propose an algorithm of medical ultrasound image edge enhancement based on directional filtering. This algorithm reduces the speckle noises first, and then segments the image based on texture analysis to determine the strong edges, the weak edges and the reverberation artifacts. Finally, based on directional filtering, the algorithm calculates the local directions of strong edge and weak edge, designs directional filters, and adaptively selects the directional filter according to the direction of image for filtering, as well as enhances the image edges. Experimental results show that the original medical ultrasound image is effectively enhanced, and the edge details are preserved. This algorithm can effectively extract edge-enhanced regions and achieves satisfied enhanced effect on tissue boundaries as well. At the same time, the reverberation artifacts of the image are kept unchanged.%由于斑点噪声的存在,传统的边缘增强算子常常会强化超声图像的斑点噪声,降低对比分辨率.针对这种情况,提出基于方向滤波的超声图像边缘增强算法.算法首先进行斑点去噪,然后基于纹理分析进行图像分割,判定图像的强边缘、弱边缘及反射伪影,最后以方向滤波为基础,计算强边缘和弱边缘的局部方向,设计方向滤波器,根据图像的方向自适应选择滤波器进行滤波,增强图像边缘.实验结果表明,原超声医学图像得到有效增强,边缘细节得以保留.该算法能有效地提取需边缘增强区域,获得满意的组织边缘增强效果,同时图像的反射伪影部分保持不变.