A Hazy Image Database with Analysis of the Frequency Magnitude

被引：4

作者：

Wang, Shuhang ^{[1
]}

Tian, Yu ^{[2
]}

Pu, Tian ^{[3
]}

Wang, Patrick ^{[4
]}

Perner, Petra ^{[5
]}

机构：

[1] Harvard Med Sch, Schepens Eye Res Inst, Mass Eye & Ear, Boston, MA 02114 USA

[2] Beihang Univ, Sch Comp Sci & Engn, Beijing 100083, Peoples R China

[3] Univ Elect Sci & Technol China, Sch Optoelect Informat, Chengdu 610054, Sichuan, Peoples R China

[4] Northeastern Univ, Coll CCIS, Boston, MA 02115 USA

[5] Inst Comp Vis & Appl Comp Sci, Leipzig, Germany

来源：

INTERNATIONAL JOURNAL OF PATTERN RECOGNITION AND ARTIFICIAL INTELLIGENCE | 2018年 / 32卷 / 05期

关键词：

Image database; haze level; haze removal; natural image statistics; frequency magnitude; QUALITY ASSESSMENT; STATISTICS;

D O I：

10.1142/S0218001418540125

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Image haze removal has been extensively studied, but there has been no such an image database regarding the haze level. It is not convenient for readers to verify the assumptions or priors that are supposed to be useful for haze removal, and meanwhile, it is not fair to compare the performance of haze removal methods, which are effective for images with different haze levels. To solve this problem, we built a database consisting of more than 3464 images of different kinds of outdoor scenes. The images of the database are grouped into four classes regarding the haze level. Along with the database, we also observe a frequency magnitude prior, i.e. the frequency magnitude decreases with the increasing haze level, which can be used as a prior to develop haze removal methods. Our purpose is to help develop image haze removal methods, as well as verify existing statistical priors and discover new ones that can be used for image processing.

引用

页数：11

共 24 条

[1]

[Anonymous], 2017, JOSA A

[2] No-Reference Quality Assessment of Contrast-Distorted Images Based on Natural Scene Statistics [J].

Fang, Yuming ;

Ma, Kede ;

Wang, Zhou ;

Lin, Weisi ;

Fang, Zhijun ;

Zhai, Guangtao .

IEEE SIGNAL PROCESSING LETTERS, 2015, 22 (07) :838-842

[3] Single Image Haze Removal Using Dark Channel Prior [J].

He, Kaiming ;

Sun, Jian ;

Tang, Xiaoou .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2011, 33 (12) :2341-2353

[4] Haze Removal Using the Difference-Structure-Preservation Prior [J].

He, Linyuan ;

Zhao, Jizhong ;

Zheng, Nanning ;

Bi, Duyan .

IEEE TRANSACTIONS ON IMAGE PROCESSING, 2017, 26 (03) :1063-1075

[5]

Koschmieder H., 1925, Theorie der horizontalen Sichtweite

[6] Single image haze removal using content-adaptive dark channel and post enhancement [J].

Li, Bo ;

Wang, Shuhang ;

Zheng, Jin ;

Zheng, Liping .

IET COMPUTER VISION, 2014, 8 (02) :131-140

[7] No-Reference Quality Assessment of Deblurred Images Based on Natural Scene Statistics [J].

Li, Leida ;

Yan, Ya ;

Lu, Zhaolin ;

Wu, Jinjian ;

Gu, Ke ;

Wang, Shiqi .

IEEE ACCESS, 2017, 5 :2163-2171

[8]

Ma KD, 2015, IEEE IMAGE PROC, P3600, DOI 10.1109/ICIP.2015.7351475

[9] Efficient Image Dehazing with Boundary Constraint and Contextual Regularization [J].

Meng, Gaofeng ;

Wang, Ying ;

Duan, Jiangyong ;

Xiang, Shiming ;

Pan, Chunhong .

2013 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2013, :617-624

[10] Blind Image Quality Assessment: From Natural Scene Statistics to Perceptual Quality [J].

Moorthy, Anush Krishna ;

Bovik, Alan Conrad .

IEEE TRANSACTIONS ON IMAGE PROCESSING, 2011, 20 (12) :3350-3364

← 1 2 3 →