CS766 Project: Image Deblurring Using Dark Cahnnel Prior

Team Member: Liang Zhang (lzhang432)

Example of Deblurring
Project Proposal:
Project Proposal, Created at April 2, 2017
Project Mid-term report:
Project Mid-Term Report, Created at April 7, 2017
Project Final Presentation:
Project Final Presentation, Created at May 8, 2017
Project Final Presentation PDF version, Created at May 8, 2017

Motaviton
Image blur is often caused by camera shake when taking the photos. As mobile photes, digital cameras and GoPros are already very common in use to take photos, and camera shake are inevitalbe, a lot of images are blured. The blur images are undesirable, sometimes the user are able to delete the blur image and retake a new photo. However, oftern the time, the capture moments are difficult to reproduce (for example, photots that were token by GoPros when the user who were skiing, or photots the were token by drone, like Dajiang). At such time, removing the blur and highly desired. Deblurring to generate higher-quality images are demanded and in greate need.


Literature Review
To deblur a image, we always need to recover a blur kernel and get a shart latent image. The deblurring is a classical problem[1] and have be researched within last decade. If blur is uniform, spatially invariant, we can use B = I k + n to model the blur process (B, I, k, n represents blur image, latent imae, blur image noise). For the blur image, we have B, but we have many pairs of I and k to the same blur image B. In order to well pose the blind deblurring, assuming sparsity of image gradient are widely used[2, 3, 4]. However, based on this perior tend to favor blurry image. Ohter deblurring methods, which favors clean iamges over blurred images are developped, for example deblurring methods based on normalized sparsity prior [5], based on internal patch recurrence[6]. However the natural image models do not handle face, text, and low illumination images well. To slove the problem, dark channel prior based method was developped[7], and was proved to handle deblurring well for nature, face, text and low illumination well[8].


Image Debluring
The image deblurring process is shown as below. We have one Blur Image, and two unkown variable. There are many cominbation of Clear Image and Blur Kernel can fit the above equation. But the problem is which combination is the right choice?



Solutions
1 Dark Channel

By comparing the dark channel of blurred image and clear image, we find that the dark channel of blurred iamge are less sparse than the dark channel of sharp image
So we want to first get dark channel of blurred image and do some irrtation to get the dark channel of clear channel (cycle in the middle of the below image), and then recover the clear image.
2 Model Optimation
Based on the fact that dark channel of sharp image have more number of zero-internsity pixels, we have the following equation:
However the last term is a non-linear min operation. This term is used to measure the sparsity of dark channel. As we already have the dark channel, we can use the dark channel as a record to slove the problem:
3 Recovery the blur kernel and clear image


Applicaton and Result

Result: Check Blure Image and the Deblurred Result
Here shows some example of result
Original Blur Image Interim Result Blur Kernel Deblur Result


Original Blur Image Interim Result Blur Kernel Deblur Result


Original Blur Image Interim Result Blur Kernel Deblur Result


Original Blur Image Interim Result Blur Kernel Deblur Result


Original Blur Image Interim Result Blur Kernel Deblur Result


Original Blur Image Interim Result Blur Kernel Deblur Result


Original Blur Image Interim Result Blur Kernel Deblur Result


Original Blur Image Interim Result Blur Kernel Deblur Result


Original Blur Image Interim Result Blur Kernel Deblur Result

My Own Data Set:
Blur Image, Download from Google Clear Image with the Deblur method


Blur Image, Download from Google Clear Image with the Deblur method


Blur Image, Download from Google Clear Image with the Deblur method

Data Set with Photos taken from my mobile phone:
Blur Image, Taken by my mobile phone Clear Image with the Deblur method



Reference

[1] L.B.Lucy. An iterative technique for the rectification of observed distributions. Astronomy Journal, 79(6):745-754, 1974.
[2] T, Chan anc C.Wong. Total variation blind deconvolution.. IEEE TIP, 7(3):370-375, 1998.
[3] R.Fergus, B.Singh, A.Hertzmann, S.T.Roweis, and W.T.Freeman. Removing camera shake from a single photograph. ACM SIGGRAPH, 25(3):787-794, 2006.
[4] Y.Hacohen, E. Shechtman, and D.Lischinski. Deblurring by example using dense correspondence. In ICCV, pages2384-2391, 2013
[5] D.Krishnan, T.Tay, and R.Fergus. Blind deconvolution using a normalized sparsit measure. In CVPR, pages2657-2664, 2011
[6] T.Michaeli and M.Irani Blind deblurring using internal patch recurrence. In ECCV, pages783-798, 2014
[7] K.He, J.Sun, and X.Tang Single image haze removal using dark channel prior. In CVPR, pages1956-1963, 2009
[8] Jinshan Pan, Deqing Sun, Hanspeter Pfister, and Ming-Hsuan Yang Blind Image Deblur- ring Using Dark Channel Prior. In CVPR, 2016


Sample Code:
Sample Code
Related Publications:
Blind Image Deblurring Using Dark Channel Prior, IEEE Conference on Computer Vison and Pattern Recognition (CVPR), 2016
Last Changed: May 08, 2017