计算机系统应用  2018, Vol. 27 Issue (12): 257-261 PDF

Research on Image Dehazing of Sea-Sky Background
ZHAO Yue-Xiu, GAO Jun-Chai, LI Han-Shan
College of Electronic and Information Engineering, Xi’an Technological University, Xi’an 710021, China
Foundation item: Program of Science and Technology Bureau, Shaanxi Province (2017GY-037); Science and Technology Innovative Team Construction Plan of Northwestern Polytechnical University
Abstract: The sea-sky background image has a large area of sky and the target must be near the sea-sky-line when looking from a distance. The existing dehazing algorithms improve the sky area by weakening the processing of the sky area. This will inevitably weaken the dehazing effect near the sea-sky-line and be detrimental to the next target detection. This study proposes the method of atmospheric scattering physical model to dehaze. Firstly, based on the features of the sea-sky background image, the edge detection algorithm is used to divide the image into sky and non-sky regions. Considering the physical meaning of atmospheric light, the maximum value of the sky region is estimated as the value of atmospheric light. Secondly, the cost function is designed based on the prior information that the fog image has low contrast and the fog-free image has high contrast, and dividing the image into blocks by SLIC superpixel segmentation. The rough transmission is estimated by finding the minimum of the function in each block, and then the guided filtering is used to eliminate block effect. Finally, the fog-free image can be obtained by substituting the parameters obtained in the first two steps into atmospheric scattering model. The experimental and analysis results show that this method can achieve better dehazing effect of sea-sky background images.
Key words: sea-sky background     dehaze     atmospheric scattering model     atmospheric light     superpixel     transmission

2009年He等[6,7]提出的暗通道理论对自然图像取得了较好的处理效果, 成为了图像复原去雾方法的一个重要研究方向. 但是由于暗通道先验在天空区域会失效, 会产生过渡区域和偏色现象, 因此涌现了一些改进的方法[8,9]. 这些改进算法大都是通过弱化对天空区域的处理来达到改善天空区域的效果, 但是这样的弱化势必会导致和天空连接处的区域也被弱化, 但对于海天背景图像, 当无人艇远距离观察时, 目标恰巧处于海天交界区域处, 因此对于He的改进算法并不适用于海天背景图像去雾.

1 经典大气散射物理模型

 $I\left( x \right) = J\left( x \right)t\left( x \right) + A\left( {1 - t\left( x \right)} \right)$ (1)

 $J\left( x \right) = A - \frac{{A - I\left( x \right)}}{{t\left( x \right)}}$ (2)
2 本文方法

2.1 改进的全局大气光A的估计

2.1.1 天空区域分割

 图 1 天空区域分割

(1)将彩色图像转换为灰度图像, 因为中值滤波能保持边缘的细节, 因此采用中值滤波对图像进行预处理, 滤除图像中的噪声, 便于进行边缘提取, 如图1中(b)所示.

(2)由于天空灰度较平坦, 利用Canny边缘检测算子获得灰度图像的边缘信息, 如图1中(c)所示, 并进行膨胀和腐蚀等数学形态学操作获得二值图像, 如(d)所示.

(3)对二值图像取反并进行区域标记, 如(e)所示, 利用天空区域必在图像上方的先验知识只保留第一部分区域, 从而得到天空区域和非天空区域, 如(f)所示.

2.1.2 大气光A的估计

 $A = \max \left( {S\left( x \right)} \right)$ (3)

2.2 改进的透射率t(x)的估计

2.2.1 改进的基于代价函数的透射率粗估计

 ${D'} = \sqrt {{{\left( {\frac{{{d_c}}}{m}} \right)}^2} + {{\left( {\frac{{{d_s}}}{S}} \right)}^2}}$ (4)

 图 2 透射率估计

 ${E_{\rm{contrast}}} = -\!\!\!\!\!\! \sum\limits_{c \in r,g,b} {\sum\limits_{p \in B} {\frac{{{{\left( {{J_c}\left( p \right) - {{\mathop J\limits^ - }_c}} \right)}^2}}}{{{N_b}}}} } = -\!\!\!\!\!\! \sum\limits_{c \in r,g,b} {\sum\limits_{p \in B} {\frac{{{{\left( {{I_c}\left( p \right) - \mathop {{I_c}}\limits^ - } \right)}^2}}}{{{t^2}{N_b}}}} }$ (5)

 ${E_{\rm{loss}}} = \sum {\sum {\left\{ {{{\left( {\min \left\{ {0,{J_c}\left( p \right)} \right\}} \right)}^2} + {{\left( {\max \left\{ {0,{J_c}\left( p \right) - 255} \right\}} \right)}^2}} \right\}} }$ (6)

 $E = {E_{\rm{contrast}}} + \lambda {E_{\rm{loss}}}$ (7)

 ${t^ * }\!\!\! =\!\!\! \max \!\!\left\{ {\mathop {\min }\limits_{c \in \left\{ {r,g,b} \right\}} \mathop {\min }\limits_{p \in B} \!\!\left\{ {\frac{{{I_c}\left( p \right) - {A_c}}}{{ - {A_c}}}} \right\},\!\!\!\mathop {\max }\limits_{c \in \left\{ {r,g,b} \right\}} \mathop {\max }\limits_{p \in B} \left\{ {\frac{{{I_c}\left( p \right) - {A_c}}}{{255 - {A_c}}}} \!\!\right\}} \!\!\right\}$ (8)

2.2.2 基于引导滤波的透射率细化

 ${q_i} = {a_k}{I_i} + {b_k},\forall i \in {\omega _k}$ (9)

3 实验与分析

 图 3 去雾效果对比

4 结论与展望

 [1] Wong CY, Jiang GN, Rahman MA, et al. Histogram equalization and optimal profile compression based approach for colour image enhancement. Journal of Visual Communication and Image Representation, 2016, 38: 802-813. DOI:10.1016/j.jvcir.2016.04.019 [2] Li Y, Zhang YF, Geng AH, et al. Infrared image enhancement based on atmospheric scattering model and histogram equalization. Optics & Laser Technology, 2016, 83: 99-107. [3] Gao YY, Hu HM, Wang SH, et al. A fast image dehazing algorithm based on negative correction. Signal Processing, 2014, 103: 380-398. DOI:10.1016/j.sigpro.2014.02.016 [4] Biswas B, Roy P, Choudhuri R, et al. Microscopic image contrast and brightness enhancement using multi-scale retinex and cuckoo search algorithm. Procedia Computer Science, 2015, 70: 348-354. DOI:10.1016/j.procs.2015.10.031 [5] Lin HN, Shi ZW. Multi-scale retinex improvement for nighttime image enhancement. Optik-International Journal for Light and Electron Optics, 2014, 125(24): 7143-7148. DOI:10.1016/j.ijleo.2014.07.118 [6] He KM, Sun J, Tang XO. Single image haze removal using dark channel prior. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(12): 2341-2353. DOI:10.1109/TPAMI.2010.168 [7] He KM, Sun J, Tang XO. Guided image filtering. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35(6): 1397-1409. DOI:10.1109/TPAMI.2012.213 [8] 陈书贞, 任占广, 练秋生. 基于改进暗通道和导向滤波的单幅图像去雾算法. 自动化学报, 2016, 42(3): 455-465. [9] 张冰冰, 戴声奎, 孙万源. 基于暗原色先验模型的快速去雾算法. 中国图象图形学报, 2013, 18(2): 184-188. [10] McCartney EJ. Optics of the atmosphere: Scattering by molecules and particles. Journal of Modern Optics, 1977, 14(7): 521-521. [11] Kim JH, Jang WD, Sim JY, et al. Optimized contrast enhancement for real-time image and video dehazing. Journal of Visual Communication and Image Representation, 2013, 24(3): 410-425. DOI:10.1016/j.jvcir.2013.02.004 [12] Achanta R, Shaji A, Smith K, et al. SLIC superpixels compared to state-of-the-art superpixel methods. IEEE Transactions on Pattern analysis and Machine Intelligence, 2012, 34(11): 2274-2282. DOI:10.1109/TPAMI.2012.120