首页 >

一种快速航空图像去雾算法

王敬东1,2, 张文涛1, 王子瑞1   

  1. 1. 南京航空航天大学 自动化学院, 江苏 南京 210016;
    2. 中国航空工业集团公司 洛阳电光设备研究所, 河南 洛阳 471000
  • 收稿日期:2012-04-27 修回日期:2012-10-10 出版日期:2013-03-25 发布日期:2013-03-29
  • 通讯作者: 王敬东,Tel.: 025-84893451 E-mail: wjd1291@nuaa.edu.cn E-mail:wjd1291@nuaa.edu.cn
  • 作者简介:王敬东 男, 硕士, 副教授。主要研究方向: 机器视觉, 计算机测控。 Tel: 025-84893451 E-mail: wjd1291@nuaa.edu.cn;张文涛 男, 硕士。主要研究方向: 数字图像处理。 E-mail: shadow-fiend@163.com;王子瑞 男, 学士。主要研究方向: 数字图像处理。 Tel: 025-84893451 E-mail: wangzirui9011@163.com;许丽红 女, 硕士研究生。主要研究方向: 数字图像处理。 Tel: 025-84893451 E-mail: xlhnuaa@163.com
  • 基金资助:

    国家自然科学基金(61074161)

A Fast Aerial Image De-haze Algorithm

WANG Jingdong1,2, ZHANG Wentao1, WANG Zirui1   

  1. 1. College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China;
    2. Luoyang Electro-optic Equipment Research Institute, Aviation Industry Corporation of China, Luoyang 471000, China
  • Received:2012-04-27 Revised:2012-10-10 Online:2013-03-25 Published:2013-03-29
  • Supported by:

    National Natural Science Foundation of China (61074161)

摘要:

基于暗元先验规律的图像去雾算法去雾效果较好,但在进行介质透射率估计时会造成图像中出现方块效应,所以需要进行图像抠图处理。而图像抠图计算复杂度非常高,使得图像去雾时间消耗很大。本文提出一种快速图像去雾算法,使用Kuwahara边缘-角点保持滤波器对大气散射光进行估计,并对所采用的Kuwahara滤波器进行改进。通过增加子块的数目以及进行局部加权等,提高边缘保留效果,抑制方块效应,从而获得较为准确的介质透射率。实验结果表明本文算法时间复杂度低,图像去雾效果好。

关键词: 图像复原, 滤波器, 图像去雾, 暗元先验规律, 边缘-角点保持滤波器

Abstract:

An image matting algorithm is employed in order to alleviate the block effect of the image transmission which is generated during conducting the dark channel prior. However, the image de-haze process is very time-consuming as the matting algorithm is of huge computational complexity. A fast algorithm for image haze removal is therefore proposed in which the Kuwahara edge-and-corner preserving filter is used to estimate the atmospheric scattering light and the Kuwahara filter is adapted. By increasing the number of sub-blocks and adjusting the local weight coefficients, etc., the edge preservation is improved and the block effect is restrained, which leads to an accurate media transmission rate. The experimental result shows the algorithm has a low time complexity, while the visual effect of the image is satisfactory after the de-haze restoration.

Key words: image restoration, filter, image de-haze, dark channel prior, edge-and-corner preserving filter

中图分类号: