YE Kai,DING Yan.A Dehazing Method Combining Physical Model and Depth Map Estimation[J].Journal of Chengdu University of Information Technology,2021,36(04):390-395.[doi:10.16836/j.cnki.jcuit.2021.04.007]
一种结合物理模型和景深估算的图像去雾算法
- Title:
- A Dehazing Method Combining Physical Model and Depth Map Estimation
- 文章编号:
- 2096-1618(2021)04-0390-06
- Keywords:
- image dehazing; depth map; atmospheric light; transmission map; atmospheric scattering coefficient
- 分类号:
- TP391.4
- 文献标志码:
- A
- 摘要:
- 图像去雾是图像处理和计算机视觉领域中的一个重要课题。为解决现有去雾算法容易造成颜色失真的问题,提出一种结合物理模型和景深估算的去雾方法。首先,采用基于四叉树细分的方法估算大气光值; 然后,通过颜色衰减先验建立线性模型,根据景深获得自适应的大气散射系数和透射率; 最后,利用大气散射模型得到无雾图像。实验结果证明:在视觉效果方面,去雾算法效果明显。在客观指标方面,利用He、Meng、Zhu方法及文中方法对5幅不同图像进行去雾,文中均方误差(MSE)分别比其他3种图像去雾算法中最低的值降低了13%,78%,23%,25%,75%,峰值信噪比(PSNR)则分别提高了4%、31%、5%、8%、31%。
- Abstract:
- Image dehazing is an important topic in the field of image processing and computer visualization. An improved image dehazing method based on color attenuation prior is proposed in this paper to address the problem of the existing dehazing algorithms that are prone to color distortion. First, a method based on quad-tree subdivision is used to estimate the atmospheric light. Then, a linear model is established using color attenuation prior, and adaptive atmospheric scattering coefficients and a transmission map are obtained according to the map depth map. Finally, a haze-free image is obtained from the atmospheric scattering model. To compare the effect of the proposed method with other existing methods, five different images are dehazed with the proposed algorithm and three other methods(He, Meng and Zhu methods). In the experiments, the proposed algorithm shows better dehazing effect than the other three methods. The mean square error(MSE)of the five images with the proposed method decreased by 13%, 78%, 23%, 25%, 75%, respectively, compared to the lowest MSE of the other three methods. The peak signal-to-noise ratios(PSNR)of the five images with the proposed method increased by 4%, 31%, 5%, 8%, 31%, respectively, compared to the highest PSNR of the other three methods.
参考文献/References:
[1] 郭璠,蔡自兴,谢斌,等.图像去雾技术研究综述与展望[J].计算机应用,2010,30(9):2417-2421.
[2] Xu Y,Wen J,Fei L k,et al.Review of video and image defogging algorithms and related studies on image restoration and enhancement[J].IEEE Access,2016,4(1):165-188.
[3] 吴迪,朱青松.图像去雾的最新研究进展[J].自动化学报,2015,41(2):221-239.
[4] 徐东彬,廖庆敏.图像去雾技术研究进展[J].中国图像图形学报,2011,16(9):1561-1576.
[5] 陈丹丹,陈莉,张永新,等.修正大气耗散函数的单幅图像去雾[J].中国图像图形学报,2017,22(6):787-796.
[6] Park D,David K,Jeon C,et al.Fast single image de-hazing using characteristics of RGB channel of foggy image[J].IEICE Transactions on Information and Systems,2013(8):1793-1799.
[7] Zeng L,Dai Y Z.Single image dehazing based on combining dark channel prior and scene radiance constraint[J].Chinese Journal of Electronics,2016,25(6):1114-1120.
[8] 王平,张云峰,包芳勋.基于雾天图像降质模型的优化去雾方法[J].中国图像图形学报,2018,23(4):605-616.
[9] Zhu M Z,He Bi W,Wu Q.Single image dehazing based on dark channel prior and energy minimization[J].IEEE Signal Processing Letters,2017,25(2):174-178.
[10] Fattal R.Single image dehazing[J].ACM Transactions on Graphics,2008,27(3):1-9.
[11] He K M,Sun J,Tang X O.Single image haze removal using dark channel prior[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2011,33(12):2341-2353.
[12] He K M,Sun J,Tang X O.Guided image filtering[J].IEEE Transactions on Software Engineering,2013,35(6):1397-1409.
[13] Lee S,Yun S,Nam J,et al.A review on dark channel prior based image dehazing algorithms[J]. EURASIP Journal on Image and Video Processing,2016,1(4):1-23.
[14] Kim J,Jang W,Sim J,et al.Optimized contrast enhancement for real-time image and video dehazing[J].Journal of Visual Communications and Image Representation,2013,24(3):410-425.
[15] Meng G F,Wang Y,Duan J Y,et al.Efficient image dehazing with boundary constraint and contextual regularization[C].Proc.the IEEE International Conference on Computer Vision,Sydney,Australia,2013:617-624.
[16] Tang K T,Yang J C,Wang J.Investigating haze-relevant features in a learning framework for image dehazing[C].Proc.Computer Vision and Pattern Recognition.IEEE,2014:2995-3002.
[17] Zhu Q S,Mai J M,Shao L.A fast single image haze removal algorithm using color attenuation prior[J].IEEE Trans Image Process,2015,24(11):3522-3533.
[18] Jiang Y T,Sun C M, Zhao Y,et al.Fog density estimation and image defogging based on surrogate modeling for optical depth[J].IEEE Trans Image Process,2017,26(7):3397-3409.
[19] Raikwar S,Tapaswi S.An improved linear depth model for single image fog removal[J].Multimedia Tools and Applications,2017,77(12):1-26.
[20] Sebastian S,Ivan C,Juan-Manuel R.Single image dehazing using a multilayer perceptron[J].Journal of Electronic Imaging,2018,27(4):43022.
[21] Narasimhan S,Nayar S.Contrast restoration of weather degraded images[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2003,25(6):713-724.
[22] Narasimhan S,Nayar S.Vision and the atmosphere[J].International Journal of Computer Vision,2002,48(3):233-254.
[23] Tan R.Visibility in bad weather from a single image[C].IEEE Conference on Computer Vision and Pattern Recognition,2008:1-8.
[24] 胡雪薇,李其申.动态大气散射系数的颜色衰减先验图像去雾[J].电视技术,2017,41(S2):122-125.
[25] Preetham A,Shirley P,Smits B.A practical analytic model for daylight[C].In SIGGRAPH,1999:91-100.
[26] Berman D,Treibitz T,Avidan S.Non-local image dehazing[C].IEEE Conference on Computer Vision and Pattern Recognition,2016:1674-1682.
备注/Memo
收稿日期:2021-03-08