基于自适应全变差的B超图像快速去噪算法

Fast denoising algorithm based on adaptive total variation of B ultrasonic image

研究了B超医学图像中的斑点噪声去除问题,提出了一个保持B超医学图像细节信息的凸自适应全变差新模型,证明了新模型解的存在唯一性,并且得出了关于该模型解的框式制约。在Split Bregman算法的基础上,提出基于框式制约的保持B超医学图像细节信息的快速算法。实验结果显示,文中的新方法在去除斑点噪声的同时很好地保存了图像的细节信息,缩短了去噪时间。

The problem of the speckle noise removal on ultrasound medical images is studied. A new convex adaptive total variation model is established to maintain the detail information of B ultrasonic medical image. The existence and uniqueness of the solution of the new model is proved, and the box constraints on the solution of the model are obtained. Based on the Split Bregman algorithm, a fast algorithm is pro- posed to maintain the detail information of the medical image. Experimental results show that the new al- gorithm can remove the speckle noise and preserve the details of the image, thus reducing the denoising time.