作　　者： ; ; ; ;

机构地区： 华南理工大学理学院物理系

出　　处： 《光学技术》 2010年第6期932-935,共4页

摘　　要： 低信噪比环境下,人们运用中值滤波与维纳滤波等作为图像预处理手段获得了一定的效果,但是图像预处理严重地影响了对焦速度。在低信噪比环境下所提出的避开图像预处理的最佳自动对焦方案为:首先,采用HSI分析法把RGB彩色图像转换为亮度信号;其次,利用能量梯度算子作为图像评价函数;最后,运用爬山算法控制策略进行自动对焦。结果表明,此方案的运算时间优势非常明显,对焦准确度最佳,可满足实时对焦系统的对焦速度与对焦准确度的要求。 Image quality under low signal-noise-ratio environment can be improved by using image pre-processing methods such as median filtering and Wiener filtering,which seriously affect the automatic focusing speed. An automatic focusing method which avoids the image pre-processing in low signal-noise-ratio environment is presented. Firstly,a RGB color image is converted into luminance signal by HSI analysis; secondly,the energy gradient operator is used as image evaluation function; finally,the hill-climbing algorithm is used as automatic focusing control strategies. The operating time is shortest compared with other methods,and its focusing accuracy is the best on our platform,which satisfy the requirements of automatic focusing speed and accuracy.

关 键 词： 低信噪比 分析法 自动对焦 图像评价函数

领　　域： [自动化与计算机技术] [自动化与计算机技术]