计算机工程与应用 ›› 2015, Vol. 51 ›› Issue (8): 156-159.

• 图形图像处理 • 上一篇    下一篇

基于二次滤波的单幅图像快速去雾算法研究

刘红平1,陈明义2   

  1. 1.长沙师范学院 电子信息工程系,长沙 410100
    2.中南大学 信息与工程学院,长沙 410083
  • 出版日期:2015-04-15 发布日期:2015-04-29

Research on fast single image defogging based on secondary filter

LIU Hongping1, CHEN Mingyi2   

  1. 1.Department of Electronics Information Engineering, Changsha Normal University, Changsha 410100, China
    2.College of Information Science & Engineering, Central South University, Changsha 410083, China
  • Online:2015-04-15 Published:2015-04-29

摘要: 为改善雾、霾气候条件下采集到的图像质量,提出了一种快速的单幅图像去雾算法。该算法利用中值滤波对暗原色先验算法进行重新定义,消除halo效应;利用均值滤波对上述结果进行二次滤波,提高透射率的估算精度以及恢复图像的亮度。该算法简单,时间耗损度低。实验结果证明,该算法可有效提高场景的对比度和清晰度,且对天空色度具有很好的恢复性。

关键词: 去雾, 二次滤波, 图像增强, 暗原色先验

Abstract: A fast single image defogging method is proposed to improve the quality of the haze image. The dark channel prior is redefined by median filter to suppress halo effect. Average filter is used for the secondary filtering to improve the estimate precision of the transmittance and the brightness of the image. This method is simple, less time consumption. The experimental results show that this algorithm can effectively improve the contrast and clarity of scene, and can recover high quality sky haze-free image.

Key words: dehazing, secondary filter, image enhancement, dark channel prior