Dissertation report for image denoising using non-local mean algorithm, discussion about subproblem of noise reduction,descrption for problem in image noise

1. A
Synopsis
on
NOISE REDUCTION OF AN IMAGE FROM NON LOCAL
MEANS VALUE ALGORITHM
Submitted for dissertation
In partial fulfilment of the degree of
Master of Technology
In
Computer Science
Submitted by
ARTI SINGH
(Roll No: xyz)
Department of Computer Science and Engineering
University Institute of Engineering & Technology
Babasaheb Bhimrao Ambedkar University, (A Central University)
Vidya Vihar, Raebareli Road, Lucknow-226025,
Uttar Pradesh, India

2. Table of Contents
Topics Page Nos.
1. Main objectives 1
2. Introduction 1
2.1 Non local means 1
2.2 Structure of the Synopsis 2
3. Related Work and Literature Survey 4-5
4. Important sub-problems 5
4.1 To identify noise in image & perform Noise reduction & image deblurring5
4.2 To Compare non local means value algorithm to other algorithm 5
4.3 Improve performance non local means algorithm 5
5. Proposed Research Work and Work-plan
5.1 Formulation of research title 5
5.2 Research Design 6
5.3 Methodology 6
Above three points (5.2.1, 5.2.2 and 5.2.3) should be separately for each Sub-problem
5.1 Estimated Work-plan (Gant/Pert Chart) 8
6. Conclusion 9
7. References 9
8. Published or Communicated work 10

3. 1
Title: Noise Reduction in image using non local means value
algorithm
1. Main objectives of the research
The main objectives of this work are summarized as under:
a) To improve quality of image perform image denoising
b) To improve quality by Perform image deblurring.
c) Compare the performance of improved non local means algorithm to original
algorithm and also with other techniques.
2. Introduction
Noise reduction from an image is an important image processing task, both as a process
itself, and as a component in other processes .Other term of Noise reduction is Image
denoising. Very many ways to denoise an image or a set of data exists. The main
properties of a good image denoising model are that it will remove noise while preserving
edges.
2.1. Non local means:
Non –local means is an algorithm in image processing for image denoising unlike “local
mean” ,which the mean value of a group of pixels surrounding a target pixel to smooth the
images .Non-local means filtering takes a mean of all pixels in the image, weighted by
how similar these pixels are to the target pixel. This results in much greater post-filtering
clarity, and less loss of detail in the image compared with local mean algorithms.

4. Noise Reduction in image using non local means value algorithm
2
The goal of image denoising methods is to recover the original image from a noisy
measurement,
v(i) = u(i) + n(i),
where v(i) is the observed value, u(i) is the “true” value and n(i) is the noise perturbation
at a pixel i. The best simple way to model the effect of noise on a digital image is to add a
Gaussian white noise
This dissertation is based on to verify the characteristics and performance of non-local
means algorithm and helps in to remove noise from the image using non-local means
algorithm.
The significance of the research is that we improve the quality of image and performance
of non-local mean methods to original non-local mean method.
2.2 Structure of the Synopsis
The remaining part of this document is organized as follows. The next section, 3 outlines the
related work done in the past and presents the literature survey. Section 4 describes the
important sub-problems. Section 5 gives the research description. In section 5.1 proposed
research title is presented, section 5.2 is devoted to research methodology used and section
5.3 gives a complete as well as estimated work-plan for the remaining time span. In section 6,
an overall conclusion of the document is presented. Section 7 lists the important references,
followed by section 8 where published and communicated works are mentioned.
3. Related Work and Literature Survey
A description about the research where different methods are used for remove noise from
image using non local mean algorithm [1]. Analysing Image Denoising using Non Local
Means Algorithm: Digital image processing remains a challenging domain of programming.
All digital images contain some degree of noise. Often times this noise is introduced by the
camera when a picture is taken. Image denoising algorithms attempt to remove this noise from
the image. In this paper the method for image denoising based on the nonlocal means (NL-
means) algorithm has been implemented and results have been developed using matlab
coding. The algorithm, called nonlocal means (NLM), uses concept of Self-Similarity. Also
images taken from the digital media like digital camera and the image taken from the internet
have been compared. The image that is taken from the internet has got aligned pixel than the
image taken from digital media. Experimental results are given to demonstrate the superior
denoising performance of the NL-means denoising technique over various image denoising
benchmarks..[2]. Non-local mean value image de-noising algorithm based on self-adaption:
With analysis of the content of image blocks, image blocks from different areas will obtain
different filter parameters and search fields, which cause the similarity weights of image
blocks a more proper distribution. Experimental results showed that the new algorithm, after
de-noising, achieved an increase of peak signal to noise ratio of images and, at the same time,
reserved details of images and marginal information effectively..[3]. A non-local algorithm
for image denoising.[4]. Image Denoising and Deblurring Using Non-Local Means Algorithm

5. Noise Reduction in image using non local means value algorithm
3
in Monochrome Images: This paper presents both areas of image restoration. Image
deblurring and denoising methods are most commonly designed for removal of both
impulsive noise and Gaussian noise. Impulsive noise is a most common noise which affects
the image quality during image acquisition, transmission, reception or storage and retrieval
process in the area of image denoising.[5].Problem of denoising in Digital image processing
and solving techniques:[6]. SURVEY ON VARIOUS NOISES AND TECHNIQUES FOR
DENOISING THE COLOR IMAGE:[7]. Fast Non-Local Means (NLM) Computation with
Probabilistic Early Termination: A speed up technique for the non-local means (NLM) image
denoising algorithm based on probabilistic early termination (PET) is proposed.[8]. AN
IMPROVED NON-LOCAL DENOISING ALGORITHM: Recently, the NL Means filter has
been proposed by Buades et al. for the suppression of white Gaussian noise. This filter
exploits the repetitive character of structures in an image, unlike conventional denoising
algorithms, which typically operate in a local neighborhood. Even though the method is quite
intuitive and potentially very powerful, the PSNR and visual results are somewhat inferior to
other recent state-of-the-art non-local algorithms, like KSVD and BM-3D. In this paper, we
show that the NL Means algorithm is basically the first iteration of the Jacobi optimization
algorithm for robustly estimating the noise-free image. Based on this insight, we present
additional improvements to the NL Means algorithm and also an extension to noise reduction
of colored (correlated) noise. For white noise, PSNR results show that the proposed method is
very competitive with the BM-3D method, while the visual quality of our method is better due
to the lower presence of artifacts. For correlated noise on the other hand, we obtain a
significant improvement in denoising performance compared to recent wavelet-based
techniques.
4. Important Sub-problems
i. To identify noise in image and perform denoising in image using non-local mean
method and other efficient techniques .and also perform image deblurring
ii. Compare the performance of non local means algorithm to other algorithm
iii. Improve the non local means algorithm for noise reduction and also compare the
performance of improved non local means algorithm to original non local means
algorithm.
5. Proposed Research Work and Work-plan
5.1 Formulation of research title
On the basis of the extensive literature survey, concerns and futuristic demand of Techniques
of image Denoising , the topic for present research work is proposed as:
“Noise Reduction in image using Non local means value
algorithm”

6. Noise Reduction in image using non local means value algorithm
4
This work shall encompass all the important sub-problems as mentioned in section 2 and will
attempt to develop algorithms for the sub-problems which are noise reduction, comparision
between denoising by nl mean algorithm ,performance of non local mean algorithm..
5.2 Research Design
This research gives a brief introduction for to identify noise in image and perform denoising
in image using non-local mean method and other efficient techniques. And perform image
deburring .And also gives description for proposed improvisation in non local means
algorithm. Firstly, we identify the noise in image then perform denoising algorithms.
5.3 Methodlogy
In this dissertation image transformation based on pixel processing has been done, which
includes image denoising. the method for image denoising based on the nonlocal means (NL-
means) algorithm has been implemented and results have been developed using matlab
coding. The algorithm, called nonlocal means (NLM), uses concept of Self-Similarity.
Experimental results are given to demonstrate the superior denoising performance of the NL-
means denoising technique over various image denoising benchmarks.
he figure shows three pixels p, q1, and q2 and their respective neighborhoods. It can be seen
that the neighborhoods of pixels p and q1 are much more similar than the neighborhoods of
pixels p and q2. In fact, to the naked eye the neighborhoods of pixels p and q2 do not seem to
be similar at all. In an image adjacent pixels are most likely to have similar neighborhoods.
But, if there is a structure in the image, non-adjacent pixels will also have similar
neighborhoods. Figure 1 illustrates this idea clearly. Most of the pixels in the same column as
p will have similar neighborhoods to p’s neighborhood. In the NLM method, the denoised
value of a pixel is determined by pixels with similar neighborhoods.

7. Noise Reduction in image using non local means value algorithm
5
5.3.1 DEBLURRING ALGORITHM
Image deblurring is the exercise of processing the whole image to view it a better
representation of the section. In this process of restoring the original sharp image a
mathematical model of the blurring is used. So,

8. Noise Reduction in image using non local means value algorithm
6
5.3 Estimated Work-plan
This section presents a list of Tasks (T) and Deliverables (D) of the research work. Table 1
summarizes the work completed and will be undertaken during the first year. Table 2 and 3
provide work plan for the remaining duration of research.
5.3.1 Completed and Remaining Work-plan of the first year
TABLE 1
Tasks (T) and
Deliverables (D)
Month/ Year
7 8 9 10 11 12 1 2 3 4 5 6
2016 2017
T1
T2
D2.1
T3
D3.1
D3.2
T4
T5
T6
T7
Table 1: First Year Work-plan
List of Tasks and Deliverables-
T1: Collection of data, materials and research papers
T2: Study of the research papers, finding and developing new concept
D2.1: Writing of literature survey
T3: Research design and methodology
D3.1: Writing methodology
D3.2: Designing model and Finite State Machine
T4: Implementation using C code
T5: Defining Data structure of Common File
T6: Writing of the thesis
T7: Submission Process and Viva

9. Noise Reduction in image using non local means value algorithm
7
6. Conclusion
This dissertation gives a generalized method for image denoising. Then in depth talk about the
non-local means algorithm for removing noise from digital image was given. The based on
simulation results, obtained by Matlab . In experimental result ,improve performance of the NL-
means algorithm
7. References
[1] Deepak Raghuvanshi, Shabahat Hasan ,Mridula Agrawal “ Analysing Image Denoising using
Non Local Means Algorithm”, International Journal of Computer Applications (0975 –
8887) Volume 56– No.13, October 2012
[2] A. Buades, B. Coll, and J Morel. “A non-local algorithm for image denoising”. IEEE
International Conference on Computer Vision and Pattern Recognition, 2005
[3] A. Buades. NL-means Pseudo-Code. http://dmi.uib.es/~tomeucoll/toni/NL-
means_code.html
[4] N. Hemalatha, “Image Denoising and Deblurring Using Non-Local Means Algorithm in
Monochrome Images”, International Journal of Engineering Research and General Science
Volume 2, Issue 2, Feb-Mar 2014,ISSN 2091-2730
[5] Mohd Awais Farooque1, Jayant S.Rohankar2, “SURVEY ON VARIOUS NOISES AND
TECHNIQUES FOR DENOISING THE COLOR IMAGE”, International Journal of
Application or Innovation in Engineering & Management (IJAIEM) Volume 2, Issue 11,
November 2013 ISSN 2319 – 4847
[6] Bart Goossens, Hiêp Luong, Aleksandra Pižurica and Wilfried Philips. “AN IMPROVED
NON-LOCAL DENOISING ALGORITHM
[7] Ramanathan Vigneshy, Byung Tae Oh_ and C.-C.Jay Kuoz “Fast Non-Local Means (NLM)
Computation with Probabilistic Early Termination”,IEEE Conference paper.
[8] A. Buades, B. Coll, and J Morel. On image denoising methods. Technical Report 2004-
15, CMLA, 2004.
[9] Ke Lu , Ning He, Liang Li, “Non-Local Based denoising for medical
images,”Computational and Mathematical methods in Medical ,vol.2012,pp.7,2012.
[10] H. Takeda, S. Farsiu, and P. Milanfar, “Kernel regression for image processing and
reconstruction,” IEEE Transactions on image processing 16(2), pp. 349–366, 2007.
[11] B. Goossens, A. Piˇzurica, and W. Philips, “Removal of Correlated Noise by
Modeling Spatial Correlations and Interscale Dependencies in the Complex Wavelet
Domain,” in Proc. of IEEE International Conference on
[12] International Journal of Computer Applications (0975 – 8887) Volume 56– No.13,
October 2012
[13] A. Piˇzurica and W. Philips, “Estimating the probability of the presence of a signal of
interest in multiresolution single and multiband image denoising,” IEEE Transactions on
image processing 15(3), pp. 654–665, 2006
[14] A. Buades, B. Coll, and J. Morel. Neighborhood filters and pde’s. Technical Report
2005-04, CMLA, 2005.

10. Noise Reduction in image using non local means value algorithm
8
[15] A. Efros and T. Leung. “Texture synthesis by nonparametric sampling.”In Proc .Int.
Conf .computer Vision, volume 2, pages 1033-1038, 1999.
[16] Awate SP, Tasdizen T, Whitaker RT. Unsupervised Texture Segmentation with
Nonparametric Neighborhood Statistics. ECCV. 2006:494–507.
[17] Huang J, Mumford D. Statistics of natural images and models. ICCV. 1999:541–547.
[18] Lee A, Pedersen K, Mumford D. The nonlinear statistics of high- contrast patches in
natural images. IJCV. 2003; 54:83–103.
[19] Mahmoudi M, Sapiro G. Fast image and video denoising via nonlocal means of
similar neighborhoods.IEEE Signal Processing Letters. 2005;12(12):839–842.
[20] Portilla J, Strela V, Wainwright M, Simoncelli E. Image denoising using scale
mixtures of gaussians in the wavelet domain. IEEE Trans On Image Processing.
2003;12:1338–1351.
[21] L. Rudin and S. Osher, “Total variation based image restoration with free local
constraints,” in Proc. Of IEEE International Conference on Image Processing (ICIP), 1, pp.
31–35, Nov. 1994.
[22] [C. Tomasi and R. Manduchi, “Bilateral filtering for gray and color images,” in
Proceedings International Conference on computer vision, pp. 839–846, 1998.
[23] Mathworks. The Matlab image processing toolbox.
http://www.mathworks.com/access/helpdesk/help/toolbox/images/
[24] L. S¸endur and I. Selesnick, “Bivariate shrinkage with local variance estimation,”
IEEE Signal Processing Letters 9, pp. 438–441, 2002.
[25] J. Portilla, V. Strela, M. Wainwright, and E. Simoncelli, “Image denoising using
scale mixtures of Gaussians in the wavelet domain ,” IEEE Transactions on image
processing 12(11), pp. 1338–1351, 2003.
8. .Published Work
Published survey paper in “ SURVEY OF NOISE IN IMAGE AND EFFICIENT
TECHNIQUE FOR NOISE REDUCTION” International Journal of Science and Research
(IJSR)..This paper has been accepted by the journal to be published.
(Mr. Ram Singar Verma) ARTISINGH
Supervisor Research Scholar