光子学报 ›› 2002, Vol. 31 ›› Issue (6): 736-738.

• 图象处理 • 上一篇    下一篇

数字图象清晰度评价函数研究

李奇1, 冯华君1, 徐之海1, 边美娟1, 申溯1, 戴瑞春2   

  1. 1. 浙江大学现代光学仪器国家重点实验室, 杭州, 310027;
    2. 中兴通讯股份有限公司上海第一研究所, 上海, 200231
  • 收稿日期:2001-10-01 出版日期:2002-06-25 发布日期:2002-06-25
  • 基金资助:
    国家教委高等学校骨干教师资助计划资助项目

DIGITAL IMAGE SHARPNESS EVALUATION FUNCTION

Li Qi1, Feng Huajun1, Xu Zhihai1, Bian Meijuan1, Shen Su1, Dai Ruichun2   

  1. 1. State Key Laboratory of Modern Optical Instrumentation, Zhejiang University, Hangzhou, China 310027;
    2. Shanghai No.1. Research Institute of Zhongxing Telecommunication Corp, Shanghai, China 200231
  • Received:2001-10-01 Online:2002-06-25 Published:2002-06-25

摘要: 分析了梯度函数、频谱函数和熵函数在自动对焦技术中的作用,并研究了评价函数的性能.实验表明,梯度函数具有运算速度快的优点,适用于需要实时成象的场合;而频谱函数具有灵敏度高、曲线尖锐的特点,能够满足高精度的要求.

关键词: 图象清晰度, 评价函数, 自动对焦

Abstract: Gradient function,frequency spectrum function and entropy function are main tools for evaluating digital image sharpness.This paper discussed their application in autofocus technology and studied feature of evaluation functions.Experimental results show that gradient function′s advantage is rapid speed and it can be used in real-time process,and frequency spectrum function has good sensibility and is qualified for higher accuracy.

Key words: Image sharpness, Evaluation function, Autofocus