CHEN Yong， TAO Meifeng， GUO Hongguang
（School of Electronics and Information Engineering, Lanzhou Jiaotong University, Lanzhou 730070, China）
Abstract： The traditional single image dehazing algorithm is susceptible to the prior knowledge of hazy image and colour distortion. A new method of deep learning multi-scale convolution neural network based on HSI colour space for single image dehazing is proposed in this paper, which directly learns the mapping relationship between hazy image and corresponding clear image in colour, saturation and brightness by the designed structure of deep learning network to achieve haze removal. Firstly, the hazy image is transformed from RGB colour space to HSI colour space. Secondly, an end-to-end multi-scale full convolution neural network model is designed. The multi-scale extraction is realized by three different dehazing sub-networks: hue H, saturation S and intensity I, and the mapping relationship between hazy image and clear image is obtained by deep learning. Finally, the model was trained and tested with hazy data set. The experimental results show that this method can achieve good dehazing effect for both synthetic hazy images and real hazy images, and is superior to other contrast algorithms in subjective and objective evaluations.
Key words： image processing; image dehazing; HSI colour space; multi-scale convolution neural network
References
［1］Shu Q L, Wu C S, Zhong Q X, et al. Alternating minimization algorithm for hybrid regularized variational image dehazing. Optik-International Journal for Light and Electron Optics, 2019, 185: 943-956.
［2］Mathias A, Samiappan D. Underwater image restoration based on diffraction bounded optimization algorithm with dark channel prior. Optik-International Journal for Light and Electron Optics, 2019, 192: 162925.
［3］Tufail Z, Khurshid K, Salman A, et al. Improved dark channel prior for image defogging using RGB and YcbCr color space. IEEE Access, 2018, 6(1): 32576-32587.
［4］Zhang W D, Dong L L, Pan X P, et al. Single image defogging based on multi-channel convolutional MSRCR. IEEE Access, 2019, 7(1): 72492-72504.
［5］Nnolim U A. Single image dehazing using adaptive dynamic stochastic resonance and wavelet-based fusion. Optik-International Journal for Light and Electron Optics, 2019, 195: 163111.
［6］Yu S Y, Zhu H. Low-illumination image enhancement algorithm based on a physical lighting model. IEEE Transactions on Circuits and Systems for Video Technology, 2019, 29(1): 28-37.
［7］Guo Y H, Ke X, Ma J,et al. A pipeline neural network for low-light image enhancement. IEEE Access, 2019, 7: 13737-13744.
［8］Kim W. Image enhancement using patch-based principal energy analysis. IEEE Access, 2018, 6: 72620-72628.
［9］Pal N S, Lal D S, Kshitij S. Visibility enhancement of images degraded by hazy weather conditions using modified non-local approach. Optik-International Journal for Light and Electron Optics, 2018, 163: 99-113.
［10］Wang J B, Lu K, Xue J, et al. Single image dehazing based on the physical model and MSRCR algorithm. IEEE Transactions on Circuits and Systems for Video Technology， 2018, 28(9): 289-292.
［11］Babu P, Vayanaperumal R. Contrast enhancement using real coded genetic algorithm based modied histogram equalization for gray scale images. International Journal of Imaging Systems and Technology, 2015, 25(1): 24-32.
［12］Wang W X, Xu L. Retinex algorithm on changing scales for haze removal with depth map. International Journal of Hybrid Information Technology, 2014, 7(4): 353-364.
［13］Ru Y, Tanaka G. Proposal of multiscale retinex using illumination adjustment for digital images. Ieice Transactions on Fundamentals of Electronics Communications & Computer Sciences, 2016, 99(11): 2003-2007.
［14］Kaplan N H. Remote sensing image enhancement using hazy image model. Optik-International Journal for Light and Electron Optics, 2018, 155: 139-148.
［15］Chaudhry A M, Riaz M M, Ghafoor A. A framework for outdoor RGB image enhancement and dehazing. IEEE Geoscience and Remote Sensing Letters, 2018, 15(6): 932-936.
［16］Zhang W F, Liang J, Ju H J, et al. Study of visibility enhancement of hazy images based on dark channel prior in polarimetric imaging. Optik-International Journal for Light and Electron Optics, 2017, 130: 123-130.
［17］Liu Y, Shang J X, Pan L, et al. A unified variational model for single image dehazing. IEEE Access, 2019, 7: 15722-15736.
［18］Ju M Y, Gu Z F, Zhang D Y, Single image haze removal based on the improved atmospheric scattering model. Neurocomputing, 2017, 260: 180-191.
［19］Wang W C, Chang F L, Ji T, et al. A fast single-image dehazing method based on a physical model and gray projection. IEEE Access, 2018, 6: 5641-5653.
［20］He K M, Sun J, Tang X C. Single image haze removal using dark channel prior. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(12): 2341-2353.
［21］He K M, Sun J, Tang X O. Guided image filtering. IEEE Transactions on Software Engineering, 2013, 35(6): 1397-1409.
［22］Rasmita L, Asimananda K, Soumya R, et al. An improvement of visual perception of single image by fusion of DCP with patch processing. Optik-International Journal for Light and Electron Optics, 2019, 183: 154-162.
［23］Deepa N, Praveen S. Color image dehazing using surround filter and dark channel prior. Journal of Visual Communication and Image Representation, 2018, 50: 9-15.
［24］Meng G F, Wang Y, Duan J Y, et al. Efficient image dehazing with boundary constraint and contextual regularization. In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Seoul, 2013: 617-624.
［25］Tang K, Yang J C, Wang J. Investigating haze-relevant features in a learning framework for image dehazing. In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Columbus, OH, USA, 2014: 2995-3002.
［26］Cai B, Xu X M, Jia K, et al. DehazeNet: an end-to-end system for single image haze removal. IEEE Transactions on Image Processing, 2016, 25(11): 5187-5198.
［27］Ren W Q, Liu S, Zhang H, et al. Single image dehazing via multi-scale convolutional neural networks with holistic edges. International Journal of Computer Vision, 2020, 128(1): 240-259.
［28］Nayar S K, Narasimhan S G. Vision in bad weather. In: Proceedings of the IEEE International Conference on Computer Vision, 1999, 2: 820-827.
［29］Narasimhan S G, Nayar S K, Vision and the atmosphere. International Journal of Computer Vision, 2002, 48(3): 233-254.
［30］Li B, Ren W, Fu D, et al. Benchmarking single image dehazing and beyond. IEEE Transactions on Image Processing. 2019, 28(1): 492-505.
［31］Wang K, Wang H, Li Y F, et al. Quantitative performance evaluation for dehazing algorithms on synthetic outdoor hazy images. IEEE Access, 2018, 6: 20481-20496.
基于HSI颜色空间的深度学习单幅图像去雾方法
陈永， 陶美风， 郭红光
(兰州交通大学 电子与信息工程学院， 甘肃 兰州 730070)
摘要:针对传统的单幅图像去雾算法容易受到雾图先验知识制约导致颜色失真等问题， 本文提出了一种基于HSI颜色空间的深度学习多尺度卷积神经网络单幅图像去雾方法， 即通过设计深度学习网络结构来直接学习雾天图像与其无雾清晰图像色调、 饱和度和亮度之间的映射关系， 从而实现图像去雾。 该方法首先将有雾图像从RGB颜色空间转换到HSI颜色空间， 然后设计了一个端到端的多尺度全卷积神经网络模型， 通过色调H、 饱和度I、 强度S三个不同的去雾子网分别进行多尺度提取， 深度学习得到有雾图像与清晰图像之间的映射关系， 从而恢复出无雾图像。 实验结果表明， 本文方法对于雾天图像具有良好的去雾效果， 在主观评价和客观评价上均优于其它对比算法。
关键词:图像处理； 图像去雾； HSI颜色空间； 多尺度卷积神经网络
引用格式：CHEN Yong, TAO Meifeng，GUO Hongguang. Deeplearning method for single image dehazing based on HSI colour space. Journal of Measurement Science and Instrumentation, 2021, 12（4）： 423-432. DOI： 10.3969／j.issn.1674-8042.2021.04.006
[full text view]
