Research On Two-dimensional Wavelet Transform Profilometry

The wavelet transform profilometry is one of the common 3D surface measurement technology based on fringe projection. It is widely used in many fields due to its ability to reconstruct 3D surface with one fringe pattern only. The 2D wavelet transform introduces direction selection analysis based on 1D wavelet transform, which improve the noise suppression ability of the wavelet transform profilometry. The analysis speed and accuracy of the fringe pattern will directly influence the speed and noise suppression ability of wavelet transform profilometry. This paper will respectively from the two-dimensional wavelet transform, the wavelet ridge extraction and phase unwrapping three aspects optimization algorithm in data analysis, the main content of research including:The selection of 2-D wavelet transform parameter is discussed in this paper. The scale factors which satisfy the geometric sequence and the rotation angles which satisfy the arithmetic progression are employed to proceed the 2D continous wavelet transform which will greatly improve the calculation speed.A new wavelet ridge extraction method for two-dimensional wavelet transform profilometry based on a new evaluation function is proposed. The function is established by using module information of the wavelet transform coefficients, fringe instantaneous frequency and local fringe direction to improve the noise suppression capability, and a fast dynamic optimization algorithm is also designed to improve the speed of wavelet ridge extraction.The phase unwrapping method is studied and the method based on the quality guiding method is employed which will further improve the noise suppression capability.Comprehensive analysis of the factors that influence of the measurement speed and the accuracy of two-dimensional wavelet transform profilometry, and the conclusion is verified by simulation and experiment...