融合灰度直方图分布的自动聚焦改进算法  

An improved auto-focusing algorithm fusing the distribution of gray histogram

作　　者：冯雁[1] 梁光明[1] 朱旭东[1] 

机构地区：[1]国防科学技术大学电子科学与工程学院,湖南长沙410073

出　　处：《现代电子技术》2015年第6期108-110,114,共4页Modern Electronics Technique

基　　金：国家创新基金(202C26214300674)

摘　　要：聚焦区域选择是自动聚焦系统的一个重要模块,在分析研究中心取窗和非均匀采样取窗等传统聚焦区域选择算法局限与不足的基础上,提出结合图像灰度直方图的自动聚焦区域选择算法,通过直方图来分离目标与背景,让目标像素直接参加清晰度计算,从而达到减少运算量的目的。实验表明,此算法加快了清晰度运算,改善了聚焦效果。The selection module of focusing region is an important part of auto？focusing system. Traditional approaches for selecting focusing region includes center window method and non？uniform sampling window method. Based on the analysis of the limitations of traditional approaches,a new auto？focusing region selection method is proposed,which combining the information of image gray histogram. The gray histogram is used to separate the target from its background. Pixels of target image are directly used to calculate the definition to reduce the computation complexity. The experimental results show that the proposed algorithm can calculate the definition more effectively and improve the focusing performance markedly.

关 键 词：灰度直方图 自动聚焦系统 区域选择算法 目标分离 

分 类 号：TN919-34[电子电信—通信与信息系统] TP391.4[电子电信—信息与通信工程]