Phase extraction of an interference fringe pattern is an important part of interferometry. The phase features of interference fringe patterns are extracted by combining phase-shifting and two-dimensional continuous wavelet transform algorithms. In this method, by combining the virtues of these two algorithms and using wavelet ridge algorithm, the phase features of fringe patterns are obtained rapidly and accurately. Through MATLAB simulation platform, the experimental results show that the signal to noise ratio obtaining by the algorithm has a great improvement comparing with traditional algorithms, and the effectiveness of phase features extraction of interference fringe patterns is proved.