【摘要】 基于机器微视觉的微机电系统(MEMS)动态测试系统,提出了一种分形小波变换亚像素检测技术提取MEMS运动轨迹算法.该算法结合电耦合器件(CCD)成像机理,利用图像的分形参数进行随机分形插值对图像边缘进行重建,通过小波变换实现重建后图像亚像素精度的边缘检测.在连续光照明条件下,对MEMS平面微运动模糊图像进行检测处理,提取和分析了MEMS运动轨迹.将该方法和在频闪条件下测得的MEMS器件的平面微运动幅值的结果进行了比对分析和讨论.由实验结果可以看出,本方法有较高的测量精度,其测量绝对误差小于0.02像素.更多还原

【Abstract】 Based on machine micro-vision dynamic testing system for micro electro mechanical system(MEMS),the technique of fractal wavelet transformation was presented to exact motion tracing.Based on CCD imaging mechanism,the edge of image was reconstructed by random fractal interpolation and the subpixel edge detection was achieved by wavelet transformation.Under the condition of continual lighting,in-plane motion tracing of MEMS devices was extracted from blur image of MEMS motion and analyzed.The in-plane motion characteristic test result acquired with this technique was compared with the results under stroboscopic condition.From experimental results it can be seen the resolution has been improved,and the absolute error value of the method is less than 0.02 pixel.更多还原