Research Of Micro-surface Profilometry Based On Phase-stepping Interferometer

The principle of white light phase-stepping interferometry is systematicallyanalyzed in this dissertation. Based on Phase-stepping interferometric technique, asurface profile measurement system was designed and implemented. The prototype ofthis system was constructed. By analyzing different phase-stepping algorithms, thephase error due to inaccurate phase-shift between data frames and image-detectornonlinearity were modeled and calculated. Accordingly, the system performance wasimproved. The residues of the interferogram affected the accuracy of phaseunwrapping. Typical algorithms are summarized and compared, their advantages anddisadvantages are analyzed, a new residues-pretreatment phase-unwrapping algorithmbased on image segmentation has been put forward. The comparison with traditionalmethod shows the advantages of the new algorithm. In this system, Closed-loop controlled piezoelectric transducer (PZT) wasadopted to minimize the phase step error, and the measurement accuracy wasincreased. Mirau interference objective is improved, and high quality interferogramwas acquired. The software of this measurement system was developed, and theaccuracy and reliability of measurement of micro topography are enhanced. Based on the experimental results, the phase-unwrapping algorithm wasimproved. The results show that measuring stability of the system is 1.7nm.Major innovations of this dissertation: Based on Phase-stepping interferometric technique, a surface profile measurement system was constructed, and system model was constructed. The system was parallel to the optical pedestal. So it is easy to be fixed and adjusted. The measurement range of the objective was expanded. The optical structure is improved by directly connecting the optical filter to the CCD camera to reduce the noise level. As a result, the quality of the interferogram was improved. Typical algorithms recently published are summarized, their advantages and disadvantages are analyzed, then a new residues-pretreatment phase-unwrapping algorithm based on image segmentation was proposed. The errors accumulation of traditional algorithms was eliminated; Compared with least-squares algorithm, this method is more efficient, which is verified by the experiment results.The Mirau interference objective was optimized. The light reflected from the back surface of the reference mirror was lowered, noise background in the interference field was reduced, and then the contrast of interference patterns was obviously enhanced. The errors of phase-stepping algorithms due to Phase-step error and image-detector nonlinearity were analyzed. Correspondingly, the system was ameliorated: Closed-loop controlled piezoelectric transducer was adopted to reduce the phase step error; the power of lighting source was optimized to minimize image detector nonlinearity. The experimental results show that the measurement accuracy was increased.