DEVELOPMENT OF PHASE RETRIEVAL TECHNIQUES FOR DIGITAL FRINGE PROJECTION PROFILOMETRY

Abstract

This thesis focuses on increasing the measurement accuracy by compressing the noise while preserving the speed for digital fringe projection profilometry (FPP). In my thesis, the first work is to study and control the phase error influence on two popular absolute phase retrieval methods. In the second work, a new multi-subzone coded algorithm is proposed to increase the phase retrieval quality without the need of projecting extra fringe patterns. In the third work, novel reference-plane-based methods are proposed for high-quality absolute phase retrieval with a minimum number of fringe patterns. Finally, an effective assessment method is proposed to quantify the phase retrieval quality obtained by the FPP techniques. Extensive experiments have been conducted to demonstrate the validities and efficiencies of the proposed methods. The experimental results illustrate that the phase retrieval quality is significantly increased using the proposed methods, which leads to a higher profile measurement accuracy.