Measurement Of MEMS In-plane Motion Based On The Sub-pixel Position Matching Algorithm

Recently, MEMS technology has entered every corner in our lives, and the process of industrialization is promoted orderly, in which the measurement technology of dynamic mechanical properties of MEMS occupies a very important position. MEMS element have a small size, small range of motion and high response frequency, which is called the "two smalls, one high" characteristics. Traditional measurement methods such as scanning electron microscopy and atomic force microscopy only suit for static morphology measurement, and is far beyond measuring the dynamic mechanical properties. This paper summarizes the previous work, based on the technology principle and material selection, to build a system for the measurement of 3D dynamic mechanical properties of MEMS, and has done an experiment on measurement of micro-machine in-plane motion.In order to solve the problem of the small size and small range of motion, a micro-vision interferometer system is proposed based on Kohler illumination and the Linnik interferometer. A systematic analysis, which includes the common path lighting technology, finite and infinite distance micro-vision technology, and interferometer adjustment technology, is introduced. This chapter provides a basis for the selection of optical devices. In order to solve the problem of high frequency response, according to the time requirements of stroboscopic micro-vision and stroboscope interferometer micro-vision, a synchronous control system of dynamic measurement based on FPGA and MCU is designed. The system controls acousto-optic modulator to generate strobe laser, decides the moment when photo is taken, and drives micro translation table to realize the five step phase shifting. This paper also design the host computer software based on MFC and study the in-plane measurement algorithm. The host computer system acquires the configuration data and processes the camera images. For the measurement of the MEMS in-plane movement, the template extraction algorithm is introduced firstly, the pixel locating algorithm with standardization covariance correlation method is introduced secondly, sub-pixel locating algorithms with step correlation method and quadric surface fitting method is described thirdly, and sequential similarity detection strategies and simplex search method to accelerate the matching process is introduced finally.A self-designed 3D dynamic MEMS measurement system is adopted in the experiment of in-plane motion measurement of micro structure driven by the translation table. Theoretical analysis shows that the maximum frequency can be achieved 500 KHz using stroboscopic measurement. Using the standard covariance correlation method to verify the pixel location, and quadric surface fitting algorithm to verify the sub-pixel location. Experimental results show that the system can complete the measurement of micro structure in-plane motion. After the further calculation, the range of movement system is 155.14?m, and the measurement accuracy is 48.5nm.