[1]吴延海,潘 晨,吴 楠.改进的Otsu递归分割单幅图像去雾算法研究[J].西安科技大学学报,2017,(03):438-444.[doi:10.13800/j.cnki.xakjdxxb.2017.0320]
 WU Yan-hai,PAN Chen,WU Nan.Single image dehazing based on an improved recursive segmentation of Otsu algorithm[J].Journal of Xi'an University of Science and Technology,2017,(03):438-444.[doi:10.13800/j.cnki.xakjdxxb.2017.0320]
点击复制

改进的Otsu递归分割单幅图像去雾算法研究(/HTML)
分享到:

西安科技大学学报[ISSN:1672-9315/CN:61-1434/N]

卷:
期数:
2017年03期
页码:
438-444
栏目:
出版日期:
2017-05-30

文章信息/Info

Title:
Single image dehazing based on an improved recursive segmentation of Otsu algorithm
文章编号:
1672-9315(2017)03-0438-07
作者:
吴延海1潘 晨1吴 楠2
1.西安科技大学 通信与信息工程学院,陕西 西安 710054; 2.陕西学前师范学院 计算机与电子信息系,陕西 西安 710100
Author(s):
WU Yan-hai1PAN Chen1WU Nan2
1.College of Communication and Information Engineering,Xi'an University of Science and Technology,Xi'an 710054,China; 2.Department of Computer and Electronic Information,Shaanxi Xueqian Normal University,Xi'an 710100,China
关键词:
图像去雾 暗原色先验 Otsu递归分割 天空区域
Keywords:
Key words:haze removal dark channel prior recursive segmentation sky region
分类号:
TP 301
DOI:
10.13800/j.cnki.xakjdxxb.2017.0320
文献标志码:
A
摘要:
针对暗原色先验去雾算法对天空区域的透射率估计值偏小问题,提出一种改进的Otsu递归分割图像去雾方法。该方法基于图像分割的思想,利用改进的Otsu递归分割算法准确分离出天空区域,结合暗原色图像灰度归一化值对天空区域透射率估计值进行修正,优化后的透射率更加接近实际值,最后通过大气散射模型得到复原图像。实验结果表明:去雾后的图像保持真实色彩,区域交叉边界处复原自然。文中算法有效解决了天空区域出现色偏和光晕问题,拓宽了暗原色先验算法的适用范围,对户外雾天图像的去雾具有良好的应用前景。
Abstract:
Abstract:In order to solve the problem of the low estimated transmission value in the sky region by the theory of dark channel prior,this paper proposed a haze removel method based on an improved recursive segmentation derived from Otsu algorithm.Under the thought of image segmentation,the method is accurate to separate sky region via the improved recursive segmentation of Otsu rule,which is combined with the normalized gray values of the dark channel map to amend the estimation for the transmission in the sky region.It optimizes the transmission to get closer to the actual values,then the haze-free image can be recovered by the atmosphere scattering model.Experimental results show that the recovered image preserves the true color and exhibits natural restoration between the boundary region.By effectively dealing with the color distortion and halos appearing in the sky region,our method broadens the application of the dark channel prior in a certain extent,which has a great prospect for processing outdoor foggy image.

参考文献/References:

[1] Tan R T.Visibility in bad weather from a single image[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2008,20(4):40-42.
[2] Fattal R.Single image dehazing[J].ACM Transactions on Graphics,2008,27(3):1-9.
[3] Tarel J P,Hautiere N.fast visibilty restoration from a single color or gray level image[C]//Computer Vision,2009 IEEE 12th International Conference on.IEEE,2009:2 201-2 208.
[4] He K,Sun J,Tang X.Single image haze removal using dark channel prior[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2010,33(12):2 341-2 353.
[5] 邢晓敏,刘 威.雾天交通场景中单幅图像去雾[J].中国图象图形学报,2016,21(11):1 440-1 447. XING Xiao-min,LIU Wei.Haze removal for single traffic image[J].Journal of Image and Graphics,2016,21(11):1 440-1 447.
[6]吴延海,张 婧,陈 康.改进的基于均值滤波的单幅图像去雾算法研究[J].西安科技大学学报,2016,36(4):583-588. WU Yan-hai,ZHANG Jing,CHEN Kang.Improved defogging algorithm of single image based on mean filter[J].Journal of Xi'an University of Science and Technology,2016,36(4):583-588.
[7]Wang G,Ren G,Jiang L,et al.Single image dehazing algorithm based on sky region segmentation[J].Information Technology Journal,2013,12(6):1 168-1 175.
[8]Otsu N.A Threshold selection method from gray-level histograms[J].IEEE Transactions on Systems Man and Cybernetics,2007,9(1):62-66.
[9]蔡燕柳,贾振红.基于改进的Otsu准则的递归图像分割算法[J].激光杂志,2008,29(4):40-42. CAI Yan-liu,JIA Zhen-hong.A recursive image segmentation algorithm based on the modified Otsu's rule[J].Laser Journal,2008,29(4):40-42.
[10]郭 昕.统计区域合并的彩色图像分割算法[J].西安科技大学学报,2015,35(3):392-396. GUO Xin.Color image segmentation method of statistical region merging[J].Journal of Xi'an University of Science and Technology,2015,35(3):392-396.
[11]He K,Sun J,Tang X.Guided image filtering[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2013,35(6):1 397-1 409.
[12]Kil T H,Sang H L,Cho N I.Single image dehazing based on reliability map of dark channel prior[C]//IEEE International Conference on Image Processing,2013:882-885.
[13]Zhu Y B,Liu J M,Hao Y G.An single image dehazing algorithm using sky detection and segmentation[C]//Image and Signal Processing(CISP),2014 7th International Congress on IEEE,2015:248-252.
[14]Wang J B,He N,Zhang L L,et al.Single image dehazing with a physical model and dark channel prior[J].Neurocomputing,2015,149(PB):718-728.

相似文献/References:

[1]吴延海,张 婧,陈 康.改进的基于均值滤波的单幅图像去雾算法研究[J].西安科技大学学报,2016,(04):583.[doi:10.13800/j.cnki.xakjdxxb.2016.0421]
 WU Yan-hai,ZHANG Jing,CHEN Kang.Improved defogging algorithm of single image based on mean filter[J].Journal of Xi'an University of Science and Technology,2016,(03):583.[doi:10.13800/j.cnki.xakjdxxb.2016.0421]

备注/Memo

备注/Memo:
基金项目:陕西省自然科学基金(2015JQ6221,2016JQ6064); 陕西学前师范学院科研项目(2015YBKJ032) 通讯作者:吴延海(1957-),男,山东荷泽人,教授,E-mail:wyh7388@163.com
更新日期/Last Update: 1900-01-01