Author Login Editor-in-Chief Peer Review Editor Work Office Work

Computer Engineering

Previous Articles     Next Articles

Autofocus Algorithm Based on Adjacent Pixel Difference and NRSS

CHEN Hao,CHEN Jian,YE Qingzhou,CAI Zhiming   

  1. (College of Information Science and Engineering,Fujian University of Technology,Fuzhou 350108,China)
  • Received:2014-08-01 Online:2015-09-15 Published:2015-09-15

基于相邻像素差与NRSS的自动对焦算法

陈浩,陈健,叶轻舟,蔡志明   

  1. (福建工程学院信息科学与工程学院,福州 350108)
  • 作者简介:陈浩(1992-),男,学士,主研方向:图像处理;陈健,讲师、博士研究生;叶轻舟,副教授、硕士;蔡志明,讲师、博士研究生。
  • 基金资助:
    福建省科技重大专项基金资助项目(2013HZ0001-4);福建省教育厅科技基金资助项目(JK2013029)。

Abstract: Traditional non-blurred image quality evaluation algorithms for real-time performance and effectiveness are not ideal for the problems existing in improving the No-reference Structural Sharpness(NRSS) algorithm.Combined with Adjacent Pixel Difference(APD),this paper proposes an autofocus-presented no reference image clarity-evaluation algorithm.It uses nearest neighbor resampling method of original image preprocessing,and computes NRSS and APD Sharpness(APDS) to the original image,uses weighted summation,and gets the final sharpness.Simulation results show that,compared with APD and NRSS algorithm,this algorithm is faster,and its evaluation result has higher consistency with subjective evaluation results.

Key words: autofocus, clarity-evaluation function, Image Quality Assessment(IQA), mean filtering, No-reference Structural Sharpness(NRSS)algorithm

摘要: 针对传统无参考模糊图像质量评价算法实时性和有效性较差的问题,在改进无参考结构清晰度(NRSS)算法的基础上,结合相邻像素差,提出一种应用于自动对焦的无参考图像清晰度评价算法。利用最邻近重采样法对原始图像进行预处理,分别计算原始图像的NRSS和相邻像素差清晰度(APDS),并将2个清晰度加权求和得到图像最终的清晰度。仿真实验结果表明,与APDS和NRSS算法相比,该算法运算速度更快,且评价结果与主观评价结果具有更高的一致性。

关键词: 自动对焦, 清晰度评价函数, 图像质量评价, 均值滤波, 无参考结构清晰度算法

CLC Number: