Phase Shifting Interferometry Algorithms And Experimental Study Of Surface Morphology Measurements

In this thesis, the author comprehensively described the research results on the theory and algorithms of phase-stepping interferometry. The study is concentrated on the theory of new phase-stepping algorithms and phase-unwrapping algorithms, and the experiments carried out to confirm and support the theory. By making some improvements to a general interferometry microscope, we calculated the phase distribution over the object surface and realized whole field restore to the phase inconstancies. The newly developed algorithms are applied to the software successfully, and the accuracy and reliability of measurement of micro-topography are highly increased. The experimental results which support the conclusion are presented. In phase-stepping algorithms, six phase-stepping algorithms which are insensitive to disturbance and error are put forward and improved: (1) Phase-stepping algorithm immune to linear reference phase error; (2) Six-frame algorithm insensitive to the reference phase and signal second humorous wave; (3) Phase-stepping algorithm insensitive to linear phase error and signal second, third, fourth humorous wave; (4) Design of phase-stepping algorithm insensitive to background modulate; (5) Algorithm insensitive to linear refenrence phase error and high humorous wave; (6) Least-square fitting phase-stepping algorithm using Lissajous figures technique. Experimental results show that the above algorithms have the characteristics of nonsensitivity to error sources. Among the algorithms the repeative measurement accuracy of algorithm (3) is less than 0.3 nm. Comparing with the conventional phase-stepping algorithms, the repeative measurement accuracy of new algorithm is advanced to more than 5 times. The noises, low modulation, error often give rise to the incorrect results of phase-unwrapping in the practical interferogram. To reduce this phenomenon, five effective and practical phase-unwrapping algorithms are developed: (1) Phase-unwrapping algorithm based on the threshold of the reference phase; (2) Phase-unwrapping algorithm based on the one dimension DCT; (3) Phase-unwrapping algorithm based on the estimation of phase jump lines; (4) Phase-unwrapping algorithm using Zernike polynomial fitting; (5) Phase-unwrapping algorithm using the ideal flat fitting. Algorithm (1), algorithm (4) and algorithm (5) are proved to be effective by computer simulation. Algorithm (2) and algorithm (3) are confirmed by experimental results. Phase-wrapping problems related to the principle and the noise in practical cases are solved. The algorithms are quite effective in manipulating test surface. Differential-interference-contrast microscopy measuring system is constructed. A serious of measuring software, algorithms softwares, and data processing softwares are compiled. Some technical difficulties existing in the surface microscopy are analyzed and solved. The property and error sources of the system are analyzed in detail. The results show that measuring accuracy is O.3nm, height measuring accuracy is 3.3nm, the longitudinal resolution of the system is 4.6nm, stability of the system is 3. Snrn.