Research And Implementation Of Data Processing Techniques In The Measurement System For 3D Dynamic MEMS Characterization

The maturing designs of moving microelectromechanical systems (MEMS) make it more-and-more important to have precise measurements and high-speed visual means to characterize dynamic microstructures. With the support of the National High Technology Research and Development Program of China (863 Program), "3D Stroboscopic interferometric Visual Measurement Technique and System for Dynamic MEMS Characterization"(No. 2002AA404100), the dissertation studies data processing techniques in the measurement system for 3D dynamic MEMS characterization.The methods of image denoise are studied, a new spin filter algorithm that can overcome the influence of 2Ï€ jumps in wrapped phase map is put forward. Three schemes of interference pattern denoise are educed after studying general filter algorithm, spin filter algorithm and SUSAN filter algorithm.The advantages and disadvantages of sub-pixel step correlation algorithm based on cubic spline interpolation, sequential similarity detection algorithm (SSDA), surface fitting method and simplex search algorithm are discussed, and these methods are syncretized in the in-plane displacement measurement. Experiment results show that the in-plane displacement is measured with accuracy of 1/50 pixel.Several typical phase unwrapping algorithms are introduced, the effectiveness of the branch cut algorithm and the quality-guided path following algorithm is validated by experiments. The necessity of group unwrapping is discussed. The selection method of the unwrapping seed point based on the figure of merit is improved to avoid "residue" being selected. Combining phase unwrapping and group unwrapping, the out-of-plane displacement of micromirror is successfully measured with satisfied results.A 3D dynamic measurement software is developed. The functions and peculiarity of the software, the 3D display based on OpenGL and the computer accurate synchronous control are described. The software has been carried out in practical measurement and has achieved fine effects.Experiment results of micromirror and microgyroscope show that the above methods are effective and practical, and the measurement system works steadily and reliably. It can meet the measurement demands of 3D dynamic MEMS characterization and has promising prospect in engineering application.