[1]栾中,尚媛园,舒华忠,等.基于SLIC和SVR的单幅图像去雾算法[J].东南大学学报(自然科学版),2018,48(1):25-29.[doi:10.3969/j.issn.1001-0505.2018.01.005]
 Luan Zhong,Shang Yuanyuan,Shu Huazhong,et al.Single image dehazing algorithm based on SLIC and SVR[J].Journal of Southeast University (Natural Science Edition),2018,48(1):25-29.[doi:10.3969/j.issn.1001-0505.2018.01.005]
点击复制

基于SLIC和SVR的单幅图像去雾算法()
分享到:

《东南大学学报(自然科学版)》[ISSN:1001-0505/CN:32-1178/N]

卷:
48
期数:
2018年第1期
页码:
25-29
栏目:
计算机科学与工程
出版日期:
2018-01-20

文章信息/Info

Title:
Single image dehazing algorithm based on SLIC and SVR
作者:
栾中12尚媛园12舒华忠3周修庄2丁辉12
1首都师范大学信息工程学院, 北京 100048; 2首都师范大学成像技术北京市高精尖创新中心, 北京 100048; 3东南大学影像科学与技术实验室, 南京210096
Author(s):
Luan Zhong12 Shang Yuanyuan12 Shu Huazhong3 Zhou Xiuzhuang2 Ding Hui12
1College of Information Engineering, Capital Normal University, Beijing 100048, China
2Beijing Advanced Innovation Center for Imaging Technology, Capital Normal University, Beijing 100048, China
3Laboratory of Image Science and Technology, Southeast University, Nanjing 210096, China
关键词:
图像恢复 图像去雾 动态大气光 支持向量机 超像素分割
Keywords:
image restoration image dehazing dynamic atmospheric light support vector regression(SVR) simple linear iterative clustering(SLIC)
分类号:
TP391
DOI:
10.3969/j.issn.1001-0505.2018.01.005
摘要:
为了提升图像去雾算法的效果和适应性,提出了一种基于简单线性迭代聚类(SLIC)超像素分割和支持向量回归(SVR)的单幅图像去雾算法.首先,将雾霾图像进行超像素分割,对于每个超像素块,取其最大通道内的最大值作为局部大气光的估计,并使用引导滤波优化大气光图的边缘.然后,采集清晰户外图像并切割成块,利用大气散射模型为其加入不同程度的雾作为训练样本,提取对比度、饱和度、直方图均衡度和最小通道均值4种特征,利用支持向量回归算法训练传输参数估算模型.实验结果表明,所提算法有效地恢复了图像的对比度和颜色饱和度,同时无论是在主观视觉效果方面,还是在结构相似度和峰值信噪比等客观评价指标方面,所提算法均优于现有传统算法.
Abstract:
To improve the effect and the adaptability of the dehazing algorithm, a single image dehazing algorithm based on simple linear iterative clustering(SLIC)super-pixel segmentation and support vector regression(SVR)is proposed. First, the haze images are segmented by super-pixel. As for each super pixel block, the local atmospheric light is estimated by the maximum value of the block maximum channel.The edge of the atmospheric light map is optimized by the guided filter. Then, the clear outdoor images are collected and split into blocks. By using the atmospheric scattering model, different levels of haze are added to these blocks to obtain the training samples. Four features including the contrast, the saturation, the histogram equalization, and the mean value of minimum channel are extracted. The transmission parameter estimation model is trained by SVR. The experimental results show that the proposed algorithm achieves good restoration for contrast and saturation. This algorithm outperforms the existing algorithms in the subjective visual effect and the objective evaluation indices such as the structural similarity and the peak signal-to-noise ratio.

参考文献/References:

[1] Narasimhan S G, Nayar S K. Vision and the atmosphere[J]. International Journal of Computer Vision, 2002(3): 233-254. DOI: 10.1023/A:1016328200723.
[2] Kim J H, Jang W D, Sim J Y, et al. Optimized contrast enhancement for real-time image and video dehazing[J]. Journal of Visual Communication and Image Representation, 2013, 24(3): 410-425. DOI:10.1016/j.jvcir.2013.02.004.
[3] He K, Sun J, Tang X. Single image haze removal using dark channel prior[J]. IEEE Trans Pattern Anal Mach Intell, 2011, 33(12): 2341-2353. DOI:10.1109/TPAMI.2010.168.
[4] Zhu Q, Mai J, Shao L. A fast single image haze removal algorithm using color attenuation prior[J]. IEEE Trans Image Process, 2015, 24(11): 3522-3533. DOI:10.1109/TIP.2015.2446191.
[5] Tarel J P, Hautiere N. Fast visibility restoration from a single color or gray level image [C]//IEEE International Conference on Computer Vision. Kyoto, Japan, 2009: 2201-2208.DOI: 10.1109/ICCV.2009.5459251.
[6] Gao Y, Hu H M, Wang S, et al. A fast image dehazing algorithm based on negative correction[J]. Signal Processing, 2014, 103: 380-398. DOI:10.1016/j.sigpro.2014.02.016.
[7] Li Z, Zheng J. Edge-preserving decomposition-based single image haze removal[J]. IEEE Trans Image Process, 2015, 24(12): 5432-5441. DOI:10.1109/TIP.2015.2482903.
[8] Tang K, Yang J, Wang J. Investigating haze-relevant features in a learning framework for image dehazing[C]//IEEE International Conference on Computer Vision and Pattern Recognition. Columbus, OH, USA,2014: 2995-3012.DOI: 10.1109/CVPR.2014.383.
[9] Cai Bolun,Xu Xiangmin, Jia Kui, et al. DehazeNet: An end-to-end system for single image haze removal[J]. IEEE Trans Image Process, 2016, 25(11): 5187-5198. DOI:10.1109/TIP.2016.2598681.
[10] Ren W, Liu S, Zhang H, et al. Single image dehazing via multi-scale convolutional neural networks [C]//European Conference on Computer Vision. Amsterdam, the Netherlands, 2016:154-169. DOI:10.1007/978-3-319-46475-6_10.
[11] Achanta R, Shaji A, Smith K, et al. SLIC superpixels compared to state-of-the-art superpixel methods[J]. IEEE Trans Pattern Anal Mach Intell, 2012, 34(11): 2274-2282. DOI:10.1109/TPAMI.2012.120.
[12] He K, Sun J, Tang X. Guided image filtering[J]. IEEE Trans Pattern Anal Mach Intell, 2013, 35(6): 1397-1409. DOI:10.1109/TPAMI.2012.213.

相似文献/References:

[1]何富运,张志胜.基于马尔科夫随机场学习模型的图像模糊核估计[J].东南大学学报(自然科学版),2016,46(6):1143.[doi:10.3969/j.issn.1001-0505.2016.06.006]
 He Fuyun,Zhang Zhisheng.Image blur kernel estimation based on Markov random field learning model[J].Journal of Southeast University (Natural Science Edition),2016,46(1):1143.[doi:10.3969/j.issn.1001-0505.2016.06.006]

备注/Memo

备注/Memo:
收稿日期: 2017-06-07.
作者简介: 栾中(1988—),男,博士生;尚媛园(联系人),女,博士,教授,博士生导师,syy@bao.ac.cn.
基金项目: 国家自然科学基金资助项目(NSF61601311)、北京市优秀人才培养资助项目(2016000020124G088)、北京市长城学者资助项目(CIT&TCD20170322)、北京市建设创新型团队和高校教师职业发展资助项目(IDHT20150507)、首都师范大学青年科研创新团队资助项目.
引用本文: 栾中,尚媛园,舒华忠,等.基于SLIC和SVR的单幅图像去雾算法[J].东南大学学报(自然科学版),2018,48(1):25-29. DOI:10.3969/j.issn.1001-0505.2018.01.005.
更新日期/Last Update: 2018-01-20