Single Image Haze Removal Using Dark Channel Prior

Haze brings troubles to many computer vision/graphics applications. It reduces the visibility of the scenes and lowers the reliability of outdoor surveillance systems; it reduces the clarity of the satellite images; it also changes the colors and decreases the contrast of daily photos, which is an annoying problem to photographers. Therefore, removing haze from images is an important and widely demanded topic in computer vision and computer graphics areas. The main challenge lies in the ambiguity of the problem. Haze attenuates the light reflected from the scenes, and further blends it with some additive light in the atmosphere. The target of haze removal is to recover the reflected light (i.e., the scene colors) from the blended light. This problem is mathematically ambiguous: there are an infinite number of solutions given the blended light. How can we know which solution is true? We need to answer this question in haze removal. Ambiguity is a common challenge for many computer vision problems. In terms of mathematics, ambiguity is because the number of equations is smaller than the number of unknowns. The methods in computer vision to solve the ambiguity can roughly categorized into two strategies. The first one is to acquire more known variables, e.g., some haze removal algorithms capture multiple images of the same scene under different settings (like polarizers).But it is not easy to obtain extra images in practice. The second strategy is to impose extra constraints using some knowledge or assumptions .All the images in this thesis are best viewed in the electronic version. This way is more practical since it requires as few as only one image. To this end, we focus on single image haze removal in this thesis. The key is to find a suitable prior. Priors are important in many computer vision topics. A prior tells the algorithm "what can we know about the fact beforehand" when the fact is not directly available. In general, a prior can be some statistical/physical properties, rules, or heuristic assumptions. The performance of the algorithms is often determined by the extent to which the prior is valid. Some widely used priors in computer vision are the smoothness prior, sparsity prior, and symmetry prior. In this thesis, we develop an effective but very simple prior, called the dark channel prior, to remove haze from a single image. The dark channel prior is a statistical property of outdoor haze-free images: most patches in these images should contain pixels which are dark in at least one color channel. These dark pixels can be due to shadows, colorfulness, geometry, or other factors. This prior provides a constraint for each pixel, and thus solves the ambiguity of the problem. Combining this prior with a physical haze imaging model, we can easily recover high quality haze-free images.

[1]  H. Helmholtz Helmholtz's Treatise on Physiological Optics , 1963 .

[2]  A. Linksz Outlines of a Theory of the Light Sense. , 1965 .

[3]  A. Gilchrist,et al.  Lightness constancy through a veiling luminance. , 1983, Journal of experimental psychology. Human perception and performance.

[4]  Franklin C. Crow,et al.  Summed-area tables for texture mapping , 1984, SIGGRAPH.

[5]  P. Chavez An improved dark-object subtraction technique for atmospheric scattering correction of multispectral data , 1988 .

[6]  Jörg Weule,et al.  Non-Linear Gaussian Filters Performing Edge Preserving Diffusion , 1995, DAGM-Symposium.

[7]  Richard F. Gunst,et al.  Applied Regression Analysis , 1999, Technometrics.

[8]  E. Adelson Lightness Perception and Lightness Illusions , 1999 .

[9]  M. Gazzaniga The new cognitive neurosciences, 2nd ed. , 2000 .

[10]  Tony F. Chan,et al.  The digital TV filter and nonlinear denoising , 2001, IEEE Trans. Image Process..

[11]  Michael Elad,et al.  On the origin of the bilateral filter and ways to improve it , 2002, IEEE Trans. Image Process..

[12]  D. Kersten,et al.  Illusions, perception and Bayes , 2002, Nature Neuroscience.

[13]  Alexei A. Efros,et al.  Fast bilateral filtering for the display of high-dynamic-range images , 2002 .

[14]  Dani Lischinski,et al.  Gradient Domain High Dynamic Range Compression , 2023 .

[15]  Daniel P. Huttenlocher,et al.  Efficient Belief Propagation for Early Vision , 2004, Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004..

[16]  Dani Lischinski,et al.  Colorization using optimization , 2004, ACM Trans. Graph..

[17]  Jonathan Winawer,et al.  Image segmentation and lightness perception , 2005, Nature.

[18]  Rüdiger Westermann,et al.  RANDOM WALKS FOR INTERACTIVE ALPHA-MATTING , 2005 .

[19]  A. AlanGilchrist Seeing in Black and White , 2006 .

[20]  Timothy A. Davis,et al.  Direct Methods for Sparse Linear Systems (Fundamentals of Algorithms 2) , 2006 .

[21]  Frédo Durand,et al.  Two-scale tone management for photographic look , 2006, ACM Trans. Graph..

[22]  Richard Szeliski,et al.  Noise Estimation from a Single Image , 2006, 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06).

[23]  Dani Lischinski,et al.  Joint bilateral upsampling , 2007, ACM Trans. Graph..

[24]  Szymon Rusinkiewicz,et al.  Multiscale shape and detail enhancement from multi-light image collections , 2007, ACM Trans. Graph..

[25]  Michael F. Cohen,et al.  Fourier Analysis of the 2D Screened Poisson Equation for Gradient Domain Problems , 2008, ECCV.

[26]  Dani Lischinski,et al.  A Closed-Form Solution to Natural Image Matting , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[27]  Michael F. Cohen,et al.  Deep photo: model-based photograph enhancement and viewing , 2008, SIGGRAPH Asia '08.

[28]  Zeev Farbman,et al.  Edge-preserving decompositions for multi-scale tone and detail manipulation , 2008, ACM Trans. Graph..

[29]  Frédo Durand,et al.  Light mixture estimation for spatially varying white balance , 2008, ACM Trans. Graph..

[30]  Raanan Fattal,et al.  Single image dehazing , 2008, ACM Trans. Graph..

[31]  Marc Levoy,et al.  Gaussian KD-trees for fast high-dimensional filtering , 2009, ACM Trans. Graph..

[32]  Jian Sun,et al.  Single image haze removal using dark channel prior , 2009, CVPR.

[33]  M. Levoy,et al.  Gaussian KD-trees for fast high-dimensional filtering , 2009, SIGGRAPH 2009.

[34]  Jian Sun,et al.  Fast matting using large kernel matting Laplacian matrices , 2010, 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[35]  Jing Xiao,et al.  Importance filtering for image retargeting , 2011, CVPR 2011.

[36]  Jian Sun,et al.  A global sampling method for alpha matting , 2011, CVPR 2011.

[37]  Jian Sun,et al.  Guided Image Filtering , 2010, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[38]  Wilhelm Burger,et al.  Digital Image Processing - An Algorithmic Introduction using Java , 2008, Texts in Computer Science.