局部二值模式超声斑点降噪

Ultrasound speckle reduction using local binary pattern

斑点噪声压缩和组织结构保留对超声图像的临床诊断非常重要,因此提出一种基于改良局部二值模式的超声斑点降噪方法。该方法通过计算像素点的局部二值模式值,将超声图像归类为斑点噪声和组织结构,并用双边滤波滤除噪声。物理体膜图像和人类肝脏图像的实验结果表明该方法可以有效滤除斑点噪声和保持图像边缘。

Speckle suppression and tissue retained are very important to clinical diagnosis of ultrasound images, therefore, a ultrasound speckle algorithm based on improved local binary pattern is put forward in this paper. Local binary pattern values of pixels in ultrasound images are computed and the whole ultrasound images are classified into speckle and tissue by these values. At last, speckle is filtered outapplying bilateral filter. Ultrasound phantom testing and in vivo liver imaging show that the proposed method can effectively preserve edges and reduce speckle.