MEMS测量中一种新型的亚像素定位技术  

Research on a Novel Sub-pixel Allocation Technology in MEMS Measurement

作　　者：罗元[1] 聂俊齐[1] 张毅[1] 

机构地区：[1]重庆邮电大学光电工程学院,重庆400065

出　　处：《半导体光电》2012年第2期261-264,共4页Semiconductor Optoelectronics

基　　金：国家自然科学基金项目(51075420)

摘　　要：亚像素定位技术是实现基于视觉的高精度MEMS测量的关键。文章提出了一种混合非线性优化和离散傅里叶变换的亚像素定位方法。首先通过相位相关法估计出像素级平移参数,然后利用互功率谱的整体信息,在其邻域范围内进行上采样矩阵傅里叶变换,精确定位相位互相关谱的峰值,测得亚像素级平移参数,最后实现了图像的高精度匹配。实验表明,该算法的定位精度优于扩展的相位相关法,像素定位精度能达到0.01像素,能够满足MEMS运动图像的高精度定位,且具有更好的噪声抑制能力和计算复杂度较低的特点。The sub-pixel allocation is a key technology for achieving high-precision measurement for MEMS.A sub-pixel allocation technology based on nonlinear optimization and DFT is proposed.First,the phase correlation method is used to estimate pixel translation parameters.Then by upsampling matrix Fourier transform,the peak of the cross correlation spectrum is allocated accurately and high-precision image matching is realized.The experimental results show that the algorithm can achieve 0.01 pixel level accuracy and is superior to the extended phase correlation method.It can meet high-precision positioning of MEMS motion image and owns anti-noise performance and lower computational complexity.

关 键 词：亚像素定位 相位相关 上采样矩阵傅里叶变换 MEMS 图像配准 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]