• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

计算机工程与科学

• 论文 • 上一篇    下一篇

一种自适应分数阶偏微分图像增强模型

李伟凯1,王政霞1,蒋伟2   

  1. (1.重庆交通大学信息科学与工程学院,重庆 400074;重庆交通大学数学与统计学院,重庆 400074)
  • 收稿日期:2016-10-02 修回日期:2016-12-04 出版日期:2018-04-25 发布日期:2018-04-25
  • 基金资助:

    国家自然科学基金(61004118,11071266,11401060);重庆市教委科学技术研究项目(KJ1500501;KJ1755492);重庆市自然科学基金(cstc2015jcyjA70010);2016年重庆市研究生科研创新项目(CYS16183);2015年重庆交通大学实验教学改革与研究项目(SYJG201532)

An adaptive fractional-order partial
differential image enhancement model

LI Weikai1,WANG Zhengxia1,JIANG Wei2   

  1. (1.School of Information Science and Engineering,Chongqing Jiaotong University,Chongqing 400074;
    2.College of Mathematics and Statistics,Chongqing Jiaotong University,Chongqing 400074,China)
  • Received:2016-10-02 Revised:2016-12-04 Online:2018-04-25 Published:2018-04-25

摘要:

针对传统的自适应分数阶偏微分方程图像增强算法对图像暗区纹理区域的增强不足的缺点,考虑到人眼对光感的敏感程度不同,将亮度对视觉的影响因素考虑进传统的自适应分数阶偏微分方程图像增强算法。以梯度和灰度值为参数,建立了一种新的自适应分数阶偏微分图像增强模型。该模型改善了传统算法对暗区图像增强不足的缺点,图像增强后的平均梯度提升明显,很好地改善了图像的视觉效果。实验结果说明本算法具有一定的有效性。
 

关键词: 图像增强, 分数阶, 自适应, 偏微分方程, 暗区图像增强

Abstract:

Aiming at the shortcoming that the conventional adaptive image enhancement algorithms based on fractionalorder partial differential equations have not good enhance performance on the dark texture region, taking into account the different sensitivity of the human eye to the light sensation, the impact of brightness on vision is combined with the conventional adaptive image enhancement algorithms based on fractionalorder partial differential equations. A new adaptive fractionalorder partial differential image enhancement model is established by using the gradient and gray value as parameters, which improves the shortcoming of the previous model in enhancing the dark region image. The average gradient of the enhanced image is increased significantly and the image’s visual effect is improved superiorly, which shows that the proposed algorithm has some validity.
 
 

Key words: image enhancement, fractional order, adaptive, partial differential equations, dark space image enhancement