In this paper, we introduce a total variation based variational model for denoising wrapped phase images. Our model improves on former methods by preserving discontinuities of the phase map and enforcing the fundamental Pythagorean trigonometric identity between the real and imaginary parts of the phase map enhancing the quality of the restored phase. The existence and uniqueness of the solution of our model is proven using standard methods. Further, we provide a fast fixed point method for finding the numerical solution and prove its convergence. Experiments on both synthetic and real patterns verify our findings.
翻译:在本文中,我们介绍了一种基于全变差的变分模型,用于去噪相位图像。相比以往的方法,我们的模型保留了相位地图的不连续性,并强制执行了基本的三角恒等式,使相位图像的实部和虚部之间的关系更加准确,从而提高了恢复后相位图的质量。我们使用标准方法证明了我们模型的解的存在性和唯一性。此外,我们提供了一种快速的不动点法来找到数值解,并证明了其收敛性。通过对合成和实际模式的实验验证了我们的结果。