基于边界梯度的自动调焦评价函数  

Evaluation Function for Automatic Focusing Based on Boundary Gradients

作　　者：吴小翠[1] 郭斯羽[1] 易国[1] 

机构地区：[1]湖南大学电气与信息工程学院,湖南长沙410082

出　　处：《计算机系统应用》2010年第3期45-48,170,共5页Computer Systems & Applications

摘　　要：介绍了若干常用的评价函数,针对基于全图像数据的评价函数的不足,提出了基于边界梯度的评价函数,经过二值边界提取后,利用边界处的边缘强度来计算评价值。为克服该函数计算费时较多的缺点,提出了一个粗调-细调结合的自动调焦方法。在实际图像上的实验结果表明,提出的方法有效,相比已有评价函数准确性更高,计算速度也可接受。Commonly used evaluation functions are introduced,and an evaluation function based on boundary gradients is proposed to overcome the shortcomings of the existing evaluation functions.The evaluation value is computed by using the edge strengths of boundary pixels after the binary boundary is obtained by edge detection procedures.Since the computation of the new function takes more time,a coarse-to-fine focusing procedure is proposed to improve the overall running speed performance.Experimental results on real images show the effectiveness of the novel method.Compared with the existing evaluation functions,the new function can achieve higher accuracy with an acceptable execution speed.

关 键 词：自动调焦 边缘检测 评价函数 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] O484.41[自动化与计算机技术—计算机科学与技术]