Relative reliability-guided phase unwrapping algorithm based on region partition and application on train wheel profilometry

In this paper, a relative reliability-guided phase unwrapping algorithm based on region partition is proposed, the designing and processing of the algorithm is discussed. Firstly, the wrapped phase map in profilometry is divided into different regions to reduce the spread of noise. Secondly, the relative reliability of adjacent pixels is used to guide the path of phase unwrapping in different regions. Finally, the adjacent regions are integrated by solving the extreme value of the weighted edge continuous cost function. Simulation data and experiment on train wheel three-dimension (3-D) measurement shows that the proposed algorithm can effectively avoid the phenomenon of stretching, which caused by noise diffusion, and has a better accuracy and a better flaw detection capability than the quality map guided algorithm.