基于极坐标和蒙特卡罗估计的超声图像去噪方法

Ultrasound image despeckling using polar coordinate system and Monte Carlo approach

针对超声图像采集特性和斑点噪声分布特点,提出极坐标系下融合蒙特卡罗估计的斑点噪声抑制方法。首先对极坐标系下含噪图像进行对数变换,然后再与估计点相关的任意径向方向进行全局域采样,根据样本点与估计点空间相关性和斑点噪声分布模型构造权重因子,最后利用蒙特卡罗方法实现斑点噪声似然加权估计。实验结果表明,该算法在滤除斑点噪声的同时,更好地保持了图像细节信息。

In view of the characteristics of ultrasound image acquisition and speckle noise statistics, this paper proposes an despeckling algorithm based on polar coordinate system and Monte Carlo estimation scheme. Firstly, it took logarithmic transformation to noisy images in polar coordinate system. Then, it acquired a set of samples in any radial direction with the estimated point in global spatial domain, and constructed weighting factor according to speckle noise model and the spatial correlation between sample point and estimated point. Consequently, the algorithmused the Monte Carlo method to achieve speckle noise likelihood-weighted estimate. Experimental results demonstrate that, the presented algorithm can better remove speckle noise while preserving image structures and details.