基于改进LK光流法的MEMS面内运动分析技术

为获取微机电系统(MEMS)面内运动过程中的动态 特征,文中提出了一个MEMS面内运 动测量系统．该系统通过FPGA产生相机外触发和频闪光源两路时钟同步驱动脉冲,利用频闪 成像的原理对谐振状态下的MEMS器件进行图像采样．对图像序列进行Canny边缘检测后,引 入Shi-Tomasi角点和Lucas-Kanade光流追踪的方法进行运动分析,获取器件的幅相信息参 数 ．使用该系统对一个定制的微陀螺仪进行实验测量,实验证明该技术可实现MEMS面内运动的 亚像素精度测量,与传统的LK光流法匹配相比,能够减少由图像灰度值变化带来的测量误差 ,同时该方法具有较好的测量精度,测量重复性可达10 nm。

LK-optical-flow-based analysis of MEMS in-plane mo tion

In order to obtain the dynamic charact eristics of micro-electro-mechan ical systems (MEMS),a MEMS in-plane motion measurement system to observe and a n alysis harmonic motions of microstructure is proposed in this paper.We present a implementation of phase-shift stroboscopic imaging system that can capture im a ges of high speed motions with a conventional camera,such as charge-coupled de v ice (CCD) camera.To achieve stroboscopic imaging,two external clock synchroniz ation pulses for external triggering of the camera and strobe light are generate d by FPGA,the triggering of the camera depends on how many times the LED light source flashes to make the camera produce a clear picture.After performing Cann y edge detection on the collected in-plane motion image sequence,the Shi-Toma si corner and Lucas-Kanade optical flow tracking method are used to perform motio n analysis to obtain the relevant motion information parameters of the device.Th e experimental measurement of a customized micro gyroscope using this system sho ws the measurement technology can achieve sub-pixel precision measurement of ME M S in-plane motion.Compared with the traditional LK optical flow method,it can reduce the measurement error caused by the change of the gray value of the image .This method has good measurement accuracy and the measurement repeatability ca n reach 10nm.