一种粗精结合的图像反馈自动调焦方法  

Visual feedback autofocusing method combining fine and rough adjustment

作　　者：杨再华[1] 李玉和[1] 李庆祥[1] 郭阳宽[1] 

机构地区：[1]清华大学精密仪器及机械学系,北京100084

出　　处：《计算机工程与设计》2005年第9期2271-2273,共3页Computer Engineering and Design

基　　金：国家973重点基础研究基金项目(2003CB716201)

摘　　要：自动调焦是保证显微成像系统实时获得清晰图像的关键技术。设计了一套适用于微操纵系统的自动调焦装置,该装置借助对图像清晰度评价来自动搜索到正焦面。对图像清晰度的评价采用了Krisch边缘梯度函数和高频分量函数相结合的方法,边缘梯度函数用于大步距粗调焦,高频分量函数用于小范围内小步距精调焦。这种方法既提高了效率,又克服了噪声的干扰。通过对不同微器件的多次调焦实验验证,该方法的调焦误差最大达到6um,满足了成像系统的要求。Autofocusing is the key technique to ensure capturing clearest image in optical systems. An autofocusing device was designed for a micomanipulator. The device is by evaluating the image definition to automatically search the focal plane. To evaluatin the image definition, the Krisch gradient function and high frequency function were combined. Krisch gradient function is used to roughly adjusted and high frequency function is to be fine adjusted in narrow range. The combination of two functions has the merits such as anti-jamming and high efficiency. The maximal deviation of the experiments is 6 um within acceptable limits from a standard.

关 键 词：自动调焦 图像清晰度 散焦图像 微操纵 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]