SlideShare a Scribd company logo
1 of 96
PARALLEL PROCESSING
FOR DIGITAL IMAGE
ENHANCEMENT
Nora Youssef Fahmy
B.Sc. in Computer and Information Sciences,
Teaching assistant at Computer Science Department
Faculty of Computer and Information Sciences
Ain Shams University
Supervised By
Prof. Dr. El-Sayed M. El-Horbaty
Dr. Abeer M. Mahmoud
Cairo 2015
AGENDA
 Introduction
 Problem Definition
 Objectives
 Related Work
 Methodologies
 Conclusions & Future Work
 Publications
 References
AGENDA
 Introduction
 Problem Definition
 Objectives
 Related Work
 Methodologies
 Conclusions & Future Work
 Publications
 References
INTRODUCTION
Image
Processing
Grayscale
Image
Medical Image
Enhancement &
Restoration
Enhancement
Domains
Noises
Quality Metrics
GRAYSCALE IMAGE
 Monochrome / One color image
 No color info
 Represent the brightness of the image
 8 bits/pixel data  256 different brightness
level
MEDICAL IMAGE
 Used to create images of the human body
 Seeks to
 Reveal internal structures hidden by the skin and bones
 Diagnose and treat disease
PET CT MRI
DEGRADATION MODEL
 Noises
 Domains
 Filters
𝒈(𝒙, 𝒚)
𝒇^
(𝒙, 𝒚)
Noise
𝜼(𝒙, 𝒚)
+
DEGRADATION RESTORATION
𝒇(𝒙,𝒚)
Degradation
function
H
Restoration
Filter(s)
GAUSSIAN NOISE
 Additive
 𝑝 𝑧 =
1
2𝜋 𝛿
𝑒−(𝑧−𝑧`) 2𝛿2
 Due to:
 Poor illumination
 High temperature
SALT & PEPPER NOISE
 Additive
 𝑝 𝑧 =
𝑃𝑎 𝑓𝑜𝑟 𝑧 = 𝑎
𝑃𝑏 𝑓𝑜𝑟 𝑧 = 𝑏
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
 Due To:
 Quick transients
 Data transmission errors
IMAGE ENHANCEMENT &
RESTORATION
 Attempts to recover an image that has been distorted
by using a priori knowledge of the degradation
process.
𝒈(𝒙, 𝒚)
𝒇^
(𝒙, 𝒚)
Noise
𝜼(𝒙, 𝒚)
+
DEGRADATION RESTORATION
𝒇(𝒙,𝒚)
Degradation
function
H
Restoration
Filter(s)
ENHANCEMENT DOMAINS
 Spatial Domain
 Transform Domain (Frequency Domain)
SPATIAL DOMAIN
 𝑔(𝑥, 𝑦) = 𝑇[𝑓(𝑥, 𝑦)]
 Direct manipulation of pixels in an
image
 Covers the neighborhood operations
(Convolution)
 Filters in this domain are time
consuming
SPATIAL FILTERING
Filters
Linear
Arithmetic Geometric
Harmonic
Contra-
Harmonic
Non-Linear
Order
Statistics
Min Max
Midpoint Median
Alpha
trimmed
Adaptive
Local Noise
Reduction
Adaptive
Median
TRANSFORM DOMAIN
(FREQUENCY DOMAIN)
 𝐹(𝑢, 𝑣) = 𝑅[𝑓(𝑥, 𝑦)]
 𝑔(𝑥, 𝑦) = 𝑅−1 𝑇[𝐹(𝑢, 𝑣)]
 A representation of an image as a sum of complex
exponentials of varying magnitudes, frequencies, and
phases
FREQUENCY FILTERING
 Filters in spatial domain have their corresponding form in
frequency domain
 Convolution in the spatial domain corresponds to multiplication in
the frequency domain, and vice versa
 Wiener filter
 Gaussian removal
 De-blurring
QUALITY METRICS
 Used for dynamically monitor and adjust image quality
 MSE =
1
𝑀𝑁 𝑖
𝑀
𝑗
𝑁
[𝑔 𝑖, 𝑗 − 𝑓(𝑖, 𝑗)]2
 𝑆𝑁𝑅 = 10 𝑙𝑜𝑔10 (
𝑖
𝑀
𝑗
𝑁
𝑓(𝑖,𝑗)
𝑖
𝑀
𝑗
𝑁[𝑔 𝑖,𝑗 −𝑓(𝑖,𝑗)]2)
 𝑃𝑆𝑁𝑅 = 10 𝑙𝑜𝑔10 (
2552
𝑀𝑆𝐸
)
INTRODUCTION
Parallel
Processing
Algorithm
Features
Tools
Decomposition
Classifications
Metrics
PARALLEL ALGORITHM FEATURES
 Granularity
 Coarse-grained
 Fine-grained
 Type of parallel processing
 Explicit
 Implicit
 Synchronization
 Latency
 Scalability
DECOMPOSITION CLASSIFICATIONS
 Data Decomposition – (Fine Grained)
Data
D1 D2 D3 D4 D5 … Dn
DECOMPOSITION CLASSIFICATIONS
 Task (Functional) Decomposition – (Coarse Grained)
Code Block (Group of Functions)
Fn1
Fn2
Fn3
Fn4
Fn5
… Fnn
TOOLS
MATALAB
Client
MATLAB
Workers
parfor
METRICS
 Speed Up  𝑆 𝑃 =
𝑇𝑠
𝑇𝑝
 Efficiency  𝐸𝐹𝐹 =
𝑆 𝑃
𝑃
AGENDA
 Introduction
 Problem Definition
 Objectives
 Related Work
 Methodologies
 Conclusions & Future Work
 Publications
 References
PROBLEM DEFINITION
 Images are corrupted by multi-noise type (i.e. Gaussian, salt &
pepper) at the same time.
 There is an open demand for multi-noise removal filters.
PROBLEM DEFINITION
 Medical images have very large size
 Processing these forms take so much time to process
sequentially
 We have to parallelize the traditional sequential
algorithms for the sake of time performance
improvement.
AGENDA
 Introduction
 Problem Definition
 Objectives
 Related Work
 Methodologies
 Conclusions & Future Work
 Publications
 References
OBJECTIVES
1. Study the literature work and implement Gaussian de-noising
experiment as a case study.
OBJECTIVES
2. Design & Develop sequential hybrid de- noising filter for multi-
noise removal and compare between it and the simple filters in terms
of peak signal to noise ratio (PSNR).
OBJECTIVES
3. Design & Develop a parallel hybrid de-noising filter then
compare it relative to the sequential hybrid above in terms of
time.
AGENDA
 Introduction
 Problem Definition
 Objectives
 Related Work
 Methodologies
 Conclusions & Future Work
 Publications
 References
RELATED WORK
 Single Noise Removal
 Multi-Noise Removal
 Parallel Image Processing
SINGLE NOISE REMOVAL
Year Author Contribution Noise to Remove Input Quality Criteria
2014 Monika P. and
Sukhdev S
Comparative study on
images de-noising
techniques
Salt & pepper Lena Image MSE
PSNR
2014 Rupinder Kaur,
Prabhpreet Kaur
Novel technique and
comparison
Speckle Ultrasound
image
SNR
PSNR
CoC
EPI
2014 P.Deepa and
M.Suganthi
Overview on image de-
noising techniques
Gaussian Lung CT image MSE
PSNR
2014 Sezal Khera and
Sheenam
Malhotra
Comparative study for
de-nosing techniques
Gaussian, Slat &
Pepper, Poison
and Speckle
MRI, X-Ray, CT PSNR
MULTI-NOISE REMOVAL
Year Authors Distortion Hybrid Input
Quality
Criteria
2014 Seema and
Meenakshi
G.
Speckle + Gaussian Median Filter based on DWT
via soft thresholding +
Mean absolute difference
Standard gray scale
images
PSNR
2013 Ankita D.
and et.al
Camera & object
motion blurs +
Gaussian +
Salt & pepper
Wiener + Median Gradient images PSNR
MSE
2013 Versha R.
and
Priyanka K.
Gaussian +
impulsive +
Speckle +
Possion
Curvelet transform +
Unsharp Mask filter +
Median filter
One gray & One
colored images
PSNR
2012 J U. and G
R.
Gaussian + Impulsive Haar wavelet filter +
soft thresholding technique +
center weighted median
10 DICOM images PSNR
MAE
UQI
ET
PARALLEL IMAGE PROCESSING
Year Author(s) Contribution Techniques
2015 Sharanjit Singh and
et.al
Implementation 1. Functions Extractions of Images
2. Clustering of 1
3. Histogram generation of 2
2014 Er.Paramjeet kaur
and Er.Nishi
Analysis CUDA, DicrectCompute, OpenCL
2013 Sanjay Saxena and
et.al
Implementation Algorithms
1. Image Segmentation
2. Image de-noising
3. Histogram Equalization
GAUSSIAN DE-NOISING EXPERIMENT
GAUSSIAN DE-NOISING ALGORITHMS
Filters
Linear
Arithmetic Geometric
Harmonic
Contra-
Harmonic
Non-Linear
Order
Statistics
Min Max
Midpoint Median
Alpha
trimmed
Adaptive
Local Noise
Reduction
Adaptive
Median
ENVIRONMENT SETTINGS
 Input:
 Mode: Gray scale – no color info
 Dimensions: average 512x512
 Distortion
 Gaussian Noise
 Mean = 0 &Variance = 1000
 Filter Size
 3x3, 5x5, 7X7 and 9x9
 Machine Specs:
 Processor: Intel core i5
 RAM: 4 GB
 OS: Windows 7 home edition, 64 bit
 Language: C#
 Tool:
SAMPLE RUN
Geometric Harmonic Midpoint
Alpha
Trimmed
Local Noise
Reduction
GroundTruth
Of 3X3
RESULTS
Avg. 3X3 pix Results
RESULTS
Avg. 5X5 pix Results
RESULTS
Avg. 7x7 pix
RESULTS
Avg. 9X9 pix.
AGENDA
 Introduction
 Problem Definition
 Objectives
 Related Work
 Methodologies
 Conclusions & Future Work
 Publications
 References
METHODOLOGIES
1. Serial Hybrid Algorithm
2. Parallel Hybrid Algorithm
SERIALHYBRIDALGORITHM
Fourier
Transform
Wiener
Inverse
Fourier
Transform
Adaptive
Median
Add Gaussian
Add Salt &
Pepper
Circular Blur
Image
Noisy
Image
Power
Spectrum
Restored
Image
Spatial
Domain
Fourier
Transform
Domain
PSEUDO-CODE
original = read image from file
original = convert to double
values
noisyBuf = degradImage(original)
procedure hybridFilter (noisyBuf)
adapOnly = apply adaptive
median filter
adapMedian(noisyBuf, 7 pix)
hybrid = apply wiener filter
wiener(adapOnly, original)
hybrid = apply
contrast(hybrid)
end procedure
procedure degradeImage(original)
I1 = Add circular blur to original
I2 = Add Gaussian noise to I1
I3 = Add Salt & Pepper noise to I3
return I3 as Noisy buffer
end procedure
1
PSEUDO-CODE
original = read image from file
original = convert to double
values
noisyBuf = degradImage(original)
procedure hybridFilter (noisyBuf)
adapOnly = apply adaptive
median filter
adapMedian(noisyBuf, 7 pix)
hybrid = apply wiener filter
wiener(adapOnly, original)
hybrid = apply
contrast(hybrid)
end procedure
procedure adapMedian(g, Smax)
f = g; %initial setup
alreadyProcessed = false(size(g));
for all window size k = start from 3 to Smax do:
get Min, max and median Values of window and
store it in zmin, zmax and zmed;
determine if we will use level b processing in
processByLvlB;
zB = (g > zmin) & (zmax > g);
outputZxy = processByLvlB & zB;
outputZmed = processByLvlB & ~zB;
f(outputZxy) = g(outputZxy);
f(outputZmed) = zmed(outputZmed);
alreadyProcessed = alreadyProcessed or
processUsingLevelB;
if all alreadyProcessed array has been done
break the loop
end if
end for
fill the remaining values in alreadyProcessed if
exist by the zmed
end procedure
1
2
PSEUDO-CODE
original = read image from file
original = convert to double
values
noisyBuf = degradImage(original)
procedure hybridFilter (noisyBuf)
adapOnly = apply adaptive
median filter
adapMedian(noisyBuf, 7 pix)
hybrid = apply wiener filter
wiener(adapOnly, original)
hybrid = apply
contrast(hybrid)
end procedure
procedure wiener(adapOnly, original)
F = create power spectrum of the adapOnly
sigma_u = calculate noise sigma in F
output = wiener deconvolution for F given the
original image and sigma_u
end procedure
1
2
3
PSEUDO-CODE
original = read image from file
original = convert to double
values
noisyBuf = degradImage(original)
procedure hybridFilter (noisyBuf)
adapOnly = apply adaptive
median filter
adapMedian(noisyBuf, 7 pix)
hybrid = apply wiener filter
wiener(adapOnly, original)
hybrid = apply
contrast(hybrid)
end procedure
procedure contrast(image)
stretch image’s pixels values to fit in range 0 -
255
end procedure
1
2
3
4
ENVIRONMENT SETTINGS
 Input:
 Mode: Gray scale – no color info
 Dimensions:
 512 X 512 (0.5 MB)
 128 X 128 (17 KB)
 Resolution: 300px
 Distortion
 Gaussian
 Salt & Pepper
 Circular Blur
 Machine Specs:
 Processor: Intel core i5
 RAM: 4 GB
 OS: Windows 7 home edition, 64 bit
 Tool:
RESULTS
 Average PSNR for Adaptive, Wiener and Serial Hybrid
Image a b c d e f g h I J k l AVG
Adaptive 10.04 14.15 14.2 12.06 13.3 13.7 13.43 9.5 13.29 17.12 6.88 12.45 12.6
Wiener 11.63 23.37 18.63 14.37 16.46 8.79 12.39 11.6 15.06 22.53 11.92 16.73 15.3
Hybrid 13.36 29.7 26.7 22.75 19.32 18.34 17.66 18.84 18.97 26.04 24.69 20.64 19.8
0
5
10
15
20
25
30
35
a b c d e f g h i j k l
PSNR
IMAGE
PSNR chart for full size images for proposed hybrid approach, adaptive
& wiener
Hybrid
Wiener
Adaptive
0
5
10
15
20
25
30
35
a b c d e f g h i j k l
PSNR
IMAGE
Thumbnails VS. full Sizes of PSNR values
Thumbnails
Full Sizes
0
10
20
30
40
50
60
a b c d e f g h i j k l
TIME(MINS)
IMAGE
Thumbnails VS. full Sizes time plot in minutes
Thumbnails
Full Size
PARALLEL HYBRID
ALGORITHM
 Adaptive Median takes
too much time
 Spatial domain
 Window size increases
Fourier
Transform
Wiener
Inverse
Fourier
Transform
Adaptive
Median
Add Gaussian
Add Salt & Pepper
Circular Blur
Image
Noisy
Image
Power
Spectrum
Restored
Image
Spatial
Domain
Fourier
Transform
Domain
PARALLELHYBRIDALGORITHM
Fourier
Transform
Wiener
Inverse
Fourier
Transform
Adaptive
Median
Add Gaussian
Add Salt &
Pepper
Circular Blur
Image
Noisy
Image
Power
Spectru
m
Restored
Image
Spatial
Domain
Fourier
Transform
Domain
PARALLELHYBRIDALGORITHM
Fourier
Transform
Wiener
Inverse
Fourier
Transform
Adaptive
Median
Add Gaussian
Add Salt &
Pepper
Circular Blur
Image
Noisy
Image
Power
Spectru
m
Restored
Image
Spatial
Domain
Fourier
Transform
Domain
PARALLELHYBRIDALGORITHM
Fourier
Transform
Wiener
Inverse
Fourier
Transform
Adaptive
Median
Add Gaussian
Add Salt &
Pepper
Circular Blur
Image
Noisy
Image
Power
Spectru
m
Restored
Image
Spatial
Domain
Fourier
Transform
Domain
PSEUDO-CODE
original = read image from file
original = convert to double
values
noisyBuf = degradImage(original)
procedure parallelHybridFilter
(noisyBuf)
adapOnly = apply adaptive
median filter
parAdapMedian(noisyBuf, 11
pix, workersNo)
hybrid = apply wiener filter
wiener(adapOnly, original)
hybrid = apply
contrast(hybrid)
end procedure
procedure degradeImage(original)
I1 = Add circular blur to original
I2 = Add Gaussian noise to I1
I3 = Add Salt & Pepper noise to I3
return I3 as Noisy buffer
end procedure
1
PSEUDO-CODE
original = read image from file
original = convert to double
values
noisyBuf = degradImage(original)
procedure parallelHybridFilter
(noisyBuf)
adapOnly = apply adaptive
median filter
parAdapMedian(noisyBuf, 11
pix, workersNo)
hybrid = apply wiener filter
wiener(adapOnly, original)
hybrid = apply
contrast(hybrid)
end procedure
procedure parAdapMedian(g, Smax, workersNo)
for each worker in workersNo do in Parallel
apply adaptive median filter convolution
procedure on g
end for
end procedure
1
2
PSEUDO-CODE
original = read image from file
original = convert to double
values
noisyBuf = degradImage(original)
procedure parallelHybridFilter
(noisyBuf)
adapOnly = apply adaptive
median filter
parAdapMedian(noisyBuf, 11
pix, workersNo)
hybrid = apply wiener filter
wiener(adapOnly, original)
hybrid = apply
contrast(hybrid)
end procedure
procedure wiener(adapOnly, original)
F = create power spectrum of the adapOnly
sigma_u = calculate noise sigma in F
output = wiener deconvolution for F given the
original image and sigma_u
end procedure
1
2
3
PSEUDO-CODE
original = read image from file
original = convert to double
values
noisyBuf = degradImage(original)
procedure parallelHybridFilter
(noisyBuf)
adapOnly = apply adaptive
median filter
parAdapMedian(noisyBuf, 11
pix, workersNo)
hybrid = apply wiener filter
wiener(adapOnly, original)
hybrid = apply
contrast(hybrid)
end procedure
procedure contrast(image)
stretch image’s pixels values to fit in range 0 -
255
end procedure
1
2
3
4
ENVIRONMENT SETTINGS
 Input:
 Mode: Gray scale – no color info - 500px
 Dimensions:
 1900 x 2368 (277 KB)
 3800 x 4736 (715 KB)
 15200 x 18944 (2.5 MB)
 Distortion:
 Gaussian
 Salt & Pepper
 Circular Blur
 Machine Specs:
 Processor: Intel core i5
 RAM: 6 GB
 OS: Windows 7 enterprise edition, 64 bit
 Tool
RESULTS
 Time in seconds for the 3 image sizes each of which
divided into 2,4 an 8 partitions
 Workers rang 2 - 12
Image Dimensions Serial Workers 2 4 6 8 10 12
1900 X 2368 79.5
2-Partitions 10.41 10.87 10.89 10.37 10.96 11.11
4- Partitions 12.1 11.99 12.22 12.29 12.49 12.23
8- Partitions 16.22 15.84 15.87 16.62 15.64 16.2
3800 X 4736 272.42
2- Partitions 44.67 41.59 36.14 35.12 42.86 49.28
4- Partitions 41.37 40.76 40.62 41.06 41.35 42.18
8- Partitions 54.88 54.65 56.63 56.85 54.69 55.21
7600 X 9472 1742.5
2- Partitions 640.34 863.87 856.24 652.75 914.24 690.8
4- Partitions 768 470.08 480.03 532.75 490.92 703.38
8- Partitions 847.9 538.55 769.91 505.67 656.4 611.59
0
100
200
300
400
500
600
700
800
2 4 6 8 10 12
TIME(SECONDS)
WORKER
Average time consumed in seconds for serial and parallel 2,4 and 8
partition input
Serial
2-Partition
4-Partition
8-Partition
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
2 4 6 8 10 12
SPEEDUP
WORKER
Average speed up for 2, 4 and 8 partition input
SP -2
Sp - 4
sp - 8
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
2 4 6 8 10 12
EFFICIENCY
WORKER
Efficiency for 2, 4 and 8 partition input
EFF -2
EFF - 4
EFF - 8
AGENDA
 Introduction
 Problem Definition
 Objectives
 Related Work
 Methodologies
 Conclusions & Future Work
 Publications
 References
CONCLUSIONS
 Simple Gaussian filters comparison showed for the average run 3, 5,
7 and 9 pix that
 Harmonic filter gave the max. PSNR
 Geometric filter gave the min. Time
CONCLUSIONS
 Presented serial hybrid filter gave PSNR enhancement
 33.33% VS. Wiener
 55.55% VS. Adaptive median
CONCLUSIONS
 Parallel hybrid filter gave a speed up
 3x for 2-partition
 4x for 4-partition
 3.5x for 8-partition
FUTURE WORK
 Apply the proposed methods medical imaging modalities like PET
(Colored).
 Try to corrupt the image with higher noise probabilities.
 Change the noise type to but take care that it suits the selected
adaptive median and wiener filters.
IMAGE PROCESSING
FUTURE WORK
 Try Different parallel model
 GPU
 CPU + GPU (Hybrid Parallel).
 Compare time performance with the existing methods.
 Integrate with cloud computing.
PARALLEL PROCESSING
AGENDA
 Introduction
 Problem Definition
 Objectives
 Related Work
 Methodologies
 Conclusions & Future Work
 Publications
 References
PUBLICATIONS
 Nora Youssef, Abeer M.Mahmoud and EL-
Sayed M.EL-Horbaty “Gaussian De-
Noising Techniques in Spatial Domain for
Gray Scale Medical Images”, Int. J. of
Information Technologies and Knowledge
(ITK), Vol. 8, No.3, pp.90-100, Bulgaria,
Jun. 2014.
PUBLICATIONS
 Nora Youssef, Abeer M.Mahmoud and
EL-Sayed M.EL-Horbaty “A Hybrid De-
Noising Technique for Multi-noise
Removal on Gray Scale Medical
Images”, Int. J. of Tomography and
Simulation (IJTS) IF 0.75, Vol. 28, No.2, pp
106-116, India, Mar. 2015.
PUBLICATIONS
 Nora Youssef, Abeer
M.Mahmoud and EL-
Sayed M.EL-Horbaty
“Gaussian De-Noising
Techniques” Lambert
Academic Publishing,
Germany, Apr. 2015.
PUBLICATIONS
 Nora Youssef, Abeer M.Mahmoud and
EL-Sayed M.EL-Horbaty “A Parallel Hybrid
Technique for Multi-Noise Removal from
Grayscale Medical Images” Int. J Real
Time Image Processing, Springer, Berlin,
May 2015 (Submitted).
AGENDA
 Introduction
 Problem Definition
 Objectives
 Related Work
 Methodologies
 Conclusions & Future Work
 Publications
 References
REFERENCES
 Ravi M. Rai and Urooz J. “Analysis Techniques For Eliminating Noise In Medical
Images Using Bivariate Shrinkage Method” Int. J. of Advanced Research in
Computer Engineering & Technology vol. 2, no. 10, pp. 2737 to 2740, 2013.
 Ankita D. and Archana S. “An Advanced Filter for Image Enhancement and
Restoration” J. Open Journal of Advanced Engineering Techniques OJAET vol.1,
no. 1, pp. 7-10, 2013.
 Salem Al-amri, N. V. Kalyankar and S. D. Khamitkar “A Comparative Study of
Removal Noise from Remote Sensing Image” Int. J. of Computer Science Issues,
vol. 7, no. 1, pp. 32 - 35, 2010.
 Sanjay S., Neeraj S. and Shiru S. "Image Processing Tasks using Parallel
Computing in Multi core Architecture and its Applications in Medical Imaging"
Int. J. of Advanced Research in Computer and Communication Engineering
vol. 2, no. 4, pp. 1896 to 1899, 2013
REFERENCES
 Rajeshwari S., Sharmila T. Sree “Efficient quality analysis of MRI image using
preprocessing techniques,” Proceedings of 2013 IEEE Conference on Information
and Communication Technologies (ICT), pp.: 391-396, 2013.
 Rafael C. Gonzalez and Richard E. Woods “Digital Image Processing” Person
Education,3rd Edition, 2008.
 Nikola R. and Milan T. “Improved Adaptive Median Filter for Denoising Ultrasound
Images” Conf. Advances in Computer Science pp. 196 - 174, 2012
 P.Deepa and M.Suganthi “Performance Evaluation of Various Denoising Filters for
Medical Image”, IJCSIT, Vol. 5, No.3, pp. 4205-4209 , 2014
REFERENCES
 B.Mohd. Jabarullah, Sandeep Saxena and Dr.C. Nelson Kennedy Badu“Survey
on Noise Removal in Digital Images” Vo. 6, No. 4, pp.45 -51 2012
 Gajanand G.“Algorithm for Image Processing Using Improved Median Filter and
Comparison of Mean, Median and Improved Median Filter” IJSCE, vol. 1, no.5
pp. 304 – 311, 2011.
 K.Selvanayaki, Dr. M. Karnan “CAD System for Automatic Detection of Brain
Tumor through Magnetic Resonance Image-A Review” Int. J. of Engineering
Science and Technology vol. 2, no.10, pp. 5890-5901, 2010.
 Monika P. and Sukhdev S. “Comparative Analysis of Image Denoising
Techniques” Int. J. of Computer Science & Engineering Technology (IJCSET) vol.
5 no. 02 pp. 160 - 167, 2014
THANKYOU!
DISCUSSION
APPENDICES
 Serial Code
 Parallel Code
 Data Sets
 Publications Materials
original = imread(imPath);
original = im2double(original);
noisyBuf = degradImage(original);
procedure hybrid (noisyBuf)
adapOnly = apply adaptive median filter (noisyBuf, 7);
hybrid = apply wiener filter (adapOnly, original);
hybrid = apply contrast(hybrid);
end procedure
procedure adapMedian(g, Smax)
f = g; %initial setup
alreadyProcessed = false(size(g));
for all window size k = start from 3 to Smax do:
zmin = get Min Value of window;
zmax = get Max Value of window;
zmed = get Median Value of window;
processByLvlB determine if we will use level b
processing;
zB = (g > zmin) & (zmax > g);
outputZxy = processByLvlB & zB;
outputZmed = processByLvlB & ~zB;
f(outputZxy) = g(outputZxy);
f(outputZmed) = zmed(outputZmed);
alreadyProcessed = alreadyProcessed or
processUsingLevelB;
if all alreadyProcessed array has been done
break the loop
end if
end for
fill the remaining values in alreadyProcessed if exist by
the zmed
end procedure
SERIAL CODE
PARALLEL CODE
original = imread(imPath);
original = im2double(original);
noisyBuf = degradImage(original);
procedure parallelHybrid (noisyBuf)
adapOnly = parallelAdapMedian(noisyBuf, 11, partition_i);
partition_i belongs to {2, 4, 8};
hybrid = apply Wiener function(adapOnly, original);
hybrid = apply contrast(hybrid);
end procedure
procedure parallelAdapMedian(g, Smax, parts)
[M, N] = size(g);
f = g; %initial setup
alreadyProcessed = false(size(g));
splittedImg = imgSpliter( f, M, N, parts );
for subimg = 1 into parts, numberOfWorkers_i do in parallel
numberOfWorkers_i belongs to {2, 4, 6 , 8, 10, 12}
myTemp = splittedImg(subimg);
t = alreadyProcessed;
for k = 3 to Smax do
zmin = get Min. value in subimg
zmax = get Max. value in subimg
zmed = get Median value in subimg
processByLvlB = determine if we will use level b
processing
zB = (g > zmin) & (zmax > g);
outputZxy = processByLvlB & zB;
outputZmed = processByLvlB & ~zB;
myTemp(outputZxy) = g(outputZxy);
myTemp(outputZmed) = zmed(outputZmed);
end for
splittedImg(subimg) = myTemp;
end parfor
end procedure
DATA SETS
 idoimagaing
 http://idoimaging.com/wiki/tiki-
index.php?page=Sample+Data
 3DISC
 http://www.3discimaging.com/our-products/medical-
solutions/firecr-medicalreaders/image-quality/
 Computed Tomography Emphysema Database
 http://image.diku.dk/emphysema_database/
 Some of images are available on Google Images.
PUBLICATIONS MATERIALS
GAUSSIAN DE-NOSING
TECHNIQUES IN SPATIAL
DOMAIN FOR GRAY SCALE
MEDICAL IMAGES
Published
International Journal
"Information Technologies &
Knowledge"
Volume 8
 Number 3
ISSN 1313-0455
Bulgaria
2014
Click here…
A HYBRID DE-NOISING TECHNIQUE
FOR MULTI-NOISE REMOVAL ON
GRAY SCALE MEDICAL IMAGES
Published
International Journal of
Tomography and Simulation
Volume, 28
 Number, 2
ISSN, 2319-3336
IF 0.75
India
2015
Click Here…
GAUSSIAN DE-NOSING
TECHNIQUES
Published
Book Chapter
ISBN: 978-3-659-49570-0
Lambert Academic Publishing
(LAP)
Germany
2015
Click Here…
A PARALLEL HYBRID TECHNIQUE
FOR MULTI-NOISE REMOVAL
FROM GRAYSCALE MEDICAL
IMAGES
Submitted
Int. J. Real Time Image Processing
Springer
IF: 1.1
Germany
2015
Click Here ...

More Related Content

What's hot

Image restoration recent_advances_and_applications
Image restoration recent_advances_and_applicationsImage restoration recent_advances_and_applications
Image restoration recent_advances_and_applicationsM Pardo
 
Speckle Noise Reduction in Ultrasound Images using Adaptive and Anisotropic D...
Speckle Noise Reduction in Ultrasound Images using Adaptive and Anisotropic D...Speckle Noise Reduction in Ultrasound Images using Adaptive and Anisotropic D...
Speckle Noise Reduction in Ultrasound Images using Adaptive and Anisotropic D...Md. Shohel Rana
 
Digital image processing short quesstion answers
Digital image processing short quesstion answersDigital image processing short quesstion answers
Digital image processing short quesstion answersAteeq Zada
 
Digital Image Processing Fundamental
Digital Image Processing FundamentalDigital Image Processing Fundamental
Digital Image Processing FundamentalThuong Nguyen Canh
 
image denoising technique using disctere wavelet transform
image denoising technique using disctere wavelet transformimage denoising technique using disctere wavelet transform
image denoising technique using disctere wavelet transformalishapb
 
Reduced Ordering Based Approach to Impulsive Noise Suppression in Color Images
Reduced Ordering Based Approach to Impulsive Noise Suppression in Color ImagesReduced Ordering Based Approach to Impulsive Noise Suppression in Color Images
Reduced Ordering Based Approach to Impulsive Noise Suppression in Color ImagesIDES Editor
 
Impulse noise removal in digital images
Impulse noise removal in digital imagesImpulse noise removal in digital images
Impulse noise removal in digital imagesMohan Raj
 
Speckle noise reduction from medical ultrasound images using wavelet thresh
Speckle noise reduction from medical ultrasound images using wavelet threshSpeckle noise reduction from medical ultrasound images using wavelet thresh
Speckle noise reduction from medical ultrasound images using wavelet threshIAEME Publication
 
Noise filtering
Noise filteringNoise filtering
Noise filteringAlaa Ahmed
 
An Approach for Image Deblurring: Based on Sparse Representation and Regulari...
An Approach for Image Deblurring: Based on Sparse Representation and Regulari...An Approach for Image Deblurring: Based on Sparse Representation and Regulari...
An Approach for Image Deblurring: Based on Sparse Representation and Regulari...IRJET Journal
 
Image enhancement techniques
Image enhancement techniquesImage enhancement techniques
Image enhancement techniquessakshij91
 
Paper id 28201452
Paper id 28201452Paper id 28201452
Paper id 28201452IJRAT
 
Image restoration
Image restorationImage restoration
Image restorationAzad Singh
 
Image restoration yogesh 201410048
Image restoration yogesh 201410048Image restoration yogesh 201410048
Image restoration yogesh 201410048yogesh kumar
 
Recovery of RGB Image from Its Halftoned Version based on DWT
Recovery of RGB Image from Its Halftoned Version based on DWTRecovery of RGB Image from Its Halftoned Version based on DWT
Recovery of RGB Image from Its Halftoned Version based on DWTIJCSIS Research Publications
 

What's hot (20)

Image restoration recent_advances_and_applications
Image restoration recent_advances_and_applicationsImage restoration recent_advances_and_applications
Image restoration recent_advances_and_applications
 
Speckle Noise Reduction in Ultrasound Images using Adaptive and Anisotropic D...
Speckle Noise Reduction in Ultrasound Images using Adaptive and Anisotropic D...Speckle Noise Reduction in Ultrasound Images using Adaptive and Anisotropic D...
Speckle Noise Reduction in Ultrasound Images using Adaptive and Anisotropic D...
 
DIP - Image Restoration
DIP - Image RestorationDIP - Image Restoration
DIP - Image Restoration
 
Digital image processing short quesstion answers
Digital image processing short quesstion answersDigital image processing short quesstion answers
Digital image processing short quesstion answers
 
Digital Image Processing Fundamental
Digital Image Processing FundamentalDigital Image Processing Fundamental
Digital Image Processing Fundamental
 
image denoising technique using disctere wavelet transform
image denoising technique using disctere wavelet transformimage denoising technique using disctere wavelet transform
image denoising technique using disctere wavelet transform
 
Reduced Ordering Based Approach to Impulsive Noise Suppression in Color Images
Reduced Ordering Based Approach to Impulsive Noise Suppression in Color ImagesReduced Ordering Based Approach to Impulsive Noise Suppression in Color Images
Reduced Ordering Based Approach to Impulsive Noise Suppression in Color Images
 
Impulse noise removal in digital images
Impulse noise removal in digital imagesImpulse noise removal in digital images
Impulse noise removal in digital images
 
Speckle noise reduction from medical ultrasound images using wavelet thresh
Speckle noise reduction from medical ultrasound images using wavelet threshSpeckle noise reduction from medical ultrasound images using wavelet thresh
Speckle noise reduction from medical ultrasound images using wavelet thresh
 
B045050812
B045050812B045050812
B045050812
 
Hg3512751279
Hg3512751279Hg3512751279
Hg3512751279
 
Noise filtering
Noise filteringNoise filtering
Noise filtering
 
final_project
final_projectfinal_project
final_project
 
An Approach for Image Deblurring: Based on Sparse Representation and Regulari...
An Approach for Image Deblurring: Based on Sparse Representation and Regulari...An Approach for Image Deblurring: Based on Sparse Representation and Regulari...
An Approach for Image Deblurring: Based on Sparse Representation and Regulari...
 
Df4201720724
Df4201720724Df4201720724
Df4201720724
 
Image enhancement techniques
Image enhancement techniquesImage enhancement techniques
Image enhancement techniques
 
Paper id 28201452
Paper id 28201452Paper id 28201452
Paper id 28201452
 
Image restoration
Image restorationImage restoration
Image restoration
 
Image restoration yogesh 201410048
Image restoration yogesh 201410048Image restoration yogesh 201410048
Image restoration yogesh 201410048
 
Recovery of RGB Image from Its Halftoned Version based on DWT
Recovery of RGB Image from Its Halftoned Version based on DWTRecovery of RGB Image from Its Halftoned Version based on DWT
Recovery of RGB Image from Its Halftoned Version based on DWT
 

Similar to Parallel Processing for Digital Image Enhancement

IRJET- A Novel Hybrid Image Denoising Technique based on Trilateral Filtering...
IRJET- A Novel Hybrid Image Denoising Technique based on Trilateral Filtering...IRJET- A Novel Hybrid Image Denoising Technique based on Trilateral Filtering...
IRJET- A Novel Hybrid Image Denoising Technique based on Trilateral Filtering...IRJET Journal
 
Image Restoration (Digital Image Processing)
Image Restoration (Digital Image Processing)Image Restoration (Digital Image Processing)
Image Restoration (Digital Image Processing)Kalyan Acharjya
 
A new methodology for sp noise removal in digital image processing
A new methodology for sp noise removal in digital image processing A new methodology for sp noise removal in digital image processing
A new methodology for sp noise removal in digital image processing ijfcstjournal
 
Lecture 6-2023.pdf
Lecture 6-2023.pdfLecture 6-2023.pdf
Lecture 6-2023.pdfssuserff72e4
 
Performance Comparison of Various Filters and Wavelet Transform for Image De-...
Performance Comparison of Various Filters and Wavelet Transform for Image De-...Performance Comparison of Various Filters and Wavelet Transform for Image De-...
Performance Comparison of Various Filters and Wavelet Transform for Image De-...IOSR Journals
 
Eurocon2009 Apalkov
Eurocon2009 ApalkovEurocon2009 Apalkov
Eurocon2009 ApalkovKhryashchev
 
IRJET- A Review on Various Restoration Techniques in Digital Image Processing
IRJET- A Review on Various Restoration Techniques in Digital Image ProcessingIRJET- A Review on Various Restoration Techniques in Digital Image Processing
IRJET- A Review on Various Restoration Techniques in Digital Image ProcessingIRJET Journal
 
Final presentation(image enhancement system)
Final presentation(image enhancement system)Final presentation(image enhancement system)
Final presentation(image enhancement system)Hammaad Khan
 
Survey Paper on Image Denoising Using Spatial Statistic son Pixel
Survey Paper on Image Denoising Using Spatial Statistic son PixelSurvey Paper on Image Denoising Using Spatial Statistic son Pixel
Survey Paper on Image Denoising Using Spatial Statistic son PixelIJERA Editor
 
Survey on Noise Removal in Digital Images
Survey on Noise Removal in Digital ImagesSurvey on Noise Removal in Digital Images
Survey on Noise Removal in Digital ImagesIOSR Journals
 
Speckle noise reduction using hybrid tmav based fuzzy filter
Speckle noise reduction using hybrid tmav based fuzzy filterSpeckle noise reduction using hybrid tmav based fuzzy filter
Speckle noise reduction using hybrid tmav based fuzzy filtereSAT Publishing House
 
Image De-noising and Enhancement for Salt and Pepper Noise using Genetic Algo...
Image De-noising and Enhancement for Salt and Pepper Noise using Genetic Algo...Image De-noising and Enhancement for Salt and Pepper Noise using Genetic Algo...
Image De-noising and Enhancement for Salt and Pepper Noise using Genetic Algo...IDES Editor
 
saltandpepper_noise_removal_2013
saltandpepper_noise_removal_2013saltandpepper_noise_removal_2013
saltandpepper_noise_removal_2013pranay yadav
 
Image Restoration Using Particle Filters By Improving The Scale Of Texture Wi...
Image Restoration Using Particle Filters By Improving The Scale Of Texture Wi...Image Restoration Using Particle Filters By Improving The Scale Of Texture Wi...
Image Restoration Using Particle Filters By Improving The Scale Of Texture Wi...CSCJournals
 
Performance Analysis and Optimization of Nonlinear Image Restoration Techniqu...
Performance Analysis and Optimization of Nonlinear Image Restoration Techniqu...Performance Analysis and Optimization of Nonlinear Image Restoration Techniqu...
Performance Analysis and Optimization of Nonlinear Image Restoration Techniqu...CSCJournals
 

Similar to Parallel Processing for Digital Image Enhancement (20)

W4101139143
W4101139143W4101139143
W4101139143
 
IRJET- A Novel Hybrid Image Denoising Technique based on Trilateral Filtering...
IRJET- A Novel Hybrid Image Denoising Technique based on Trilateral Filtering...IRJET- A Novel Hybrid Image Denoising Technique based on Trilateral Filtering...
IRJET- A Novel Hybrid Image Denoising Technique based on Trilateral Filtering...
 
Gg2411291135
Gg2411291135Gg2411291135
Gg2411291135
 
Image Restoration (Digital Image Processing)
Image Restoration (Digital Image Processing)Image Restoration (Digital Image Processing)
Image Restoration (Digital Image Processing)
 
A new methodology for sp noise removal in digital image processing
A new methodology for sp noise removal in digital image processing A new methodology for sp noise removal in digital image processing
A new methodology for sp noise removal in digital image processing
 
Lecture 6-2023.pdf
Lecture 6-2023.pdfLecture 6-2023.pdf
Lecture 6-2023.pdf
 
Performance Comparison of Various Filters and Wavelet Transform for Image De-...
Performance Comparison of Various Filters and Wavelet Transform for Image De-...Performance Comparison of Various Filters and Wavelet Transform for Image De-...
Performance Comparison of Various Filters and Wavelet Transform for Image De-...
 
Eurocon2009 Apalkov
Eurocon2009 ApalkovEurocon2009 Apalkov
Eurocon2009 Apalkov
 
IRJET- A Review on Various Restoration Techniques in Digital Image Processing
IRJET- A Review on Various Restoration Techniques in Digital Image ProcessingIRJET- A Review on Various Restoration Techniques in Digital Image Processing
IRJET- A Review on Various Restoration Techniques in Digital Image Processing
 
M.sc.iii sem digital image processing unit iv
M.sc.iii sem digital image processing unit ivM.sc.iii sem digital image processing unit iv
M.sc.iii sem digital image processing unit iv
 
Final presentation(image enhancement system)
Final presentation(image enhancement system)Final presentation(image enhancement system)
Final presentation(image enhancement system)
 
Unit3 dip
Unit3 dipUnit3 dip
Unit3 dip
 
Survey Paper on Image Denoising Using Spatial Statistic son Pixel
Survey Paper on Image Denoising Using Spatial Statistic son PixelSurvey Paper on Image Denoising Using Spatial Statistic son Pixel
Survey Paper on Image Denoising Using Spatial Statistic son Pixel
 
Survey on Noise Removal in Digital Images
Survey on Noise Removal in Digital ImagesSurvey on Noise Removal in Digital Images
Survey on Noise Removal in Digital Images
 
vs.pptx
vs.pptxvs.pptx
vs.pptx
 
Speckle noise reduction using hybrid tmav based fuzzy filter
Speckle noise reduction using hybrid tmav based fuzzy filterSpeckle noise reduction using hybrid tmav based fuzzy filter
Speckle noise reduction using hybrid tmav based fuzzy filter
 
Image De-noising and Enhancement for Salt and Pepper Noise using Genetic Algo...
Image De-noising and Enhancement for Salt and Pepper Noise using Genetic Algo...Image De-noising and Enhancement for Salt and Pepper Noise using Genetic Algo...
Image De-noising and Enhancement for Salt and Pepper Noise using Genetic Algo...
 
saltandpepper_noise_removal_2013
saltandpepper_noise_removal_2013saltandpepper_noise_removal_2013
saltandpepper_noise_removal_2013
 
Image Restoration Using Particle Filters By Improving The Scale Of Texture Wi...
Image Restoration Using Particle Filters By Improving The Scale Of Texture Wi...Image Restoration Using Particle Filters By Improving The Scale Of Texture Wi...
Image Restoration Using Particle Filters By Improving The Scale Of Texture Wi...
 
Performance Analysis and Optimization of Nonlinear Image Restoration Techniqu...
Performance Analysis and Optimization of Nonlinear Image Restoration Techniqu...Performance Analysis and Optimization of Nonlinear Image Restoration Techniqu...
Performance Analysis and Optimization of Nonlinear Image Restoration Techniqu...
 

Recently uploaded

ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITYISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITYKayeClaireEstoconing
 
4.18.24 Movement Legacies, Reflection, and Review.pptx
4.18.24 Movement Legacies, Reflection, and Review.pptx4.18.24 Movement Legacies, Reflection, and Review.pptx
4.18.24 Movement Legacies, Reflection, and Review.pptxmary850239
 
Like-prefer-love -hate+verb+ing & silent letters & citizenship text.pdf
Like-prefer-love -hate+verb+ing & silent letters & citizenship text.pdfLike-prefer-love -hate+verb+ing & silent letters & citizenship text.pdf
Like-prefer-love -hate+verb+ing & silent letters & citizenship text.pdfMr Bounab Samir
 
Difference Between Search & Browse Methods in Odoo 17
Difference Between Search & Browse Methods in Odoo 17Difference Between Search & Browse Methods in Odoo 17
Difference Between Search & Browse Methods in Odoo 17Celine George
 
MULTIDISCIPLINRY NATURE OF THE ENVIRONMENTAL STUDIES.pptx
MULTIDISCIPLINRY NATURE OF THE ENVIRONMENTAL STUDIES.pptxMULTIDISCIPLINRY NATURE OF THE ENVIRONMENTAL STUDIES.pptx
MULTIDISCIPLINRY NATURE OF THE ENVIRONMENTAL STUDIES.pptxAnupkumar Sharma
 
ENGLISH6-Q4-W3.pptxqurter our high choom
ENGLISH6-Q4-W3.pptxqurter our high choomENGLISH6-Q4-W3.pptxqurter our high choom
ENGLISH6-Q4-W3.pptxqurter our high choomnelietumpap1
 
Culture Uniformity or Diversity IN SOCIOLOGY.pptx
Culture Uniformity or Diversity IN SOCIOLOGY.pptxCulture Uniformity or Diversity IN SOCIOLOGY.pptx
Culture Uniformity or Diversity IN SOCIOLOGY.pptxPoojaSen20
 
Transaction Management in Database Management System
Transaction Management in Database Management SystemTransaction Management in Database Management System
Transaction Management in Database Management SystemChristalin Nelson
 
Proudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptxProudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptxthorishapillay1
 
call girls in Kamla Market (DELHI) 🔝 >༒9953330565🔝 genuine Escort Service 🔝✔️✔️
call girls in Kamla Market (DELHI) 🔝 >༒9953330565🔝 genuine Escort Service 🔝✔️✔️call girls in Kamla Market (DELHI) 🔝 >༒9953330565🔝 genuine Escort Service 🔝✔️✔️
call girls in Kamla Market (DELHI) 🔝 >༒9953330565🔝 genuine Escort Service 🔝✔️✔️9953056974 Low Rate Call Girls In Saket, Delhi NCR
 
Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)Mark Reed
 
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...Nguyen Thanh Tu Collection
 
INTRODUCTION TO CATHOLIC CHRISTOLOGY.pptx
INTRODUCTION TO CATHOLIC CHRISTOLOGY.pptxINTRODUCTION TO CATHOLIC CHRISTOLOGY.pptx
INTRODUCTION TO CATHOLIC CHRISTOLOGY.pptxHumphrey A Beña
 
ACC 2024 Chronicles. Cardiology. Exam.pdf
ACC 2024 Chronicles. Cardiology. Exam.pdfACC 2024 Chronicles. Cardiology. Exam.pdf
ACC 2024 Chronicles. Cardiology. Exam.pdfSpandanaRallapalli
 
Concurrency Control in Database Management system
Concurrency Control in Database Management systemConcurrency Control in Database Management system
Concurrency Control in Database Management systemChristalin Nelson
 
Barangay Council for the Protection of Children (BCPC) Orientation.pptx
Barangay Council for the Protection of Children (BCPC) Orientation.pptxBarangay Council for the Protection of Children (BCPC) Orientation.pptx
Barangay Council for the Protection of Children (BCPC) Orientation.pptxCarlos105
 
Virtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdf
Virtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdfVirtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdf
Virtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdfErwinPantujan2
 

Recently uploaded (20)

ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITYISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
 
4.18.24 Movement Legacies, Reflection, and Review.pptx
4.18.24 Movement Legacies, Reflection, and Review.pptx4.18.24 Movement Legacies, Reflection, and Review.pptx
4.18.24 Movement Legacies, Reflection, and Review.pptx
 
Like-prefer-love -hate+verb+ing & silent letters & citizenship text.pdf
Like-prefer-love -hate+verb+ing & silent letters & citizenship text.pdfLike-prefer-love -hate+verb+ing & silent letters & citizenship text.pdf
Like-prefer-love -hate+verb+ing & silent letters & citizenship text.pdf
 
Raw materials used in Herbal Cosmetics.pptx
Raw materials used in Herbal Cosmetics.pptxRaw materials used in Herbal Cosmetics.pptx
Raw materials used in Herbal Cosmetics.pptx
 
Difference Between Search & Browse Methods in Odoo 17
Difference Between Search & Browse Methods in Odoo 17Difference Between Search & Browse Methods in Odoo 17
Difference Between Search & Browse Methods in Odoo 17
 
MULTIDISCIPLINRY NATURE OF THE ENVIRONMENTAL STUDIES.pptx
MULTIDISCIPLINRY NATURE OF THE ENVIRONMENTAL STUDIES.pptxMULTIDISCIPLINRY NATURE OF THE ENVIRONMENTAL STUDIES.pptx
MULTIDISCIPLINRY NATURE OF THE ENVIRONMENTAL STUDIES.pptx
 
ENGLISH6-Q4-W3.pptxqurter our high choom
ENGLISH6-Q4-W3.pptxqurter our high choomENGLISH6-Q4-W3.pptxqurter our high choom
ENGLISH6-Q4-W3.pptxqurter our high choom
 
Culture Uniformity or Diversity IN SOCIOLOGY.pptx
Culture Uniformity or Diversity IN SOCIOLOGY.pptxCulture Uniformity or Diversity IN SOCIOLOGY.pptx
Culture Uniformity or Diversity IN SOCIOLOGY.pptx
 
Transaction Management in Database Management System
Transaction Management in Database Management SystemTransaction Management in Database Management System
Transaction Management in Database Management System
 
Proudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptxProudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptx
 
call girls in Kamla Market (DELHI) 🔝 >༒9953330565🔝 genuine Escort Service 🔝✔️✔️
call girls in Kamla Market (DELHI) 🔝 >༒9953330565🔝 genuine Escort Service 🔝✔️✔️call girls in Kamla Market (DELHI) 🔝 >༒9953330565🔝 genuine Escort Service 🔝✔️✔️
call girls in Kamla Market (DELHI) 🔝 >༒9953330565🔝 genuine Escort Service 🔝✔️✔️
 
Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)
 
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
 
INTRODUCTION TO CATHOLIC CHRISTOLOGY.pptx
INTRODUCTION TO CATHOLIC CHRISTOLOGY.pptxINTRODUCTION TO CATHOLIC CHRISTOLOGY.pptx
INTRODUCTION TO CATHOLIC CHRISTOLOGY.pptx
 
FINALS_OF_LEFT_ON_C'N_EL_DORADO_2024.pptx
FINALS_OF_LEFT_ON_C'N_EL_DORADO_2024.pptxFINALS_OF_LEFT_ON_C'N_EL_DORADO_2024.pptx
FINALS_OF_LEFT_ON_C'N_EL_DORADO_2024.pptx
 
ACC 2024 Chronicles. Cardiology. Exam.pdf
ACC 2024 Chronicles. Cardiology. Exam.pdfACC 2024 Chronicles. Cardiology. Exam.pdf
ACC 2024 Chronicles. Cardiology. Exam.pdf
 
Concurrency Control in Database Management system
Concurrency Control in Database Management systemConcurrency Control in Database Management system
Concurrency Control in Database Management system
 
Barangay Council for the Protection of Children (BCPC) Orientation.pptx
Barangay Council for the Protection of Children (BCPC) Orientation.pptxBarangay Council for the Protection of Children (BCPC) Orientation.pptx
Barangay Council for the Protection of Children (BCPC) Orientation.pptx
 
YOUVE_GOT_EMAIL_PRELIMS_EL_DORADO_2024.pptx
YOUVE_GOT_EMAIL_PRELIMS_EL_DORADO_2024.pptxYOUVE_GOT_EMAIL_PRELIMS_EL_DORADO_2024.pptx
YOUVE_GOT_EMAIL_PRELIMS_EL_DORADO_2024.pptx
 
Virtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdf
Virtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdfVirtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdf
Virtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdf
 

Parallel Processing for Digital Image Enhancement

  • 1. PARALLEL PROCESSING FOR DIGITAL IMAGE ENHANCEMENT Nora Youssef Fahmy B.Sc. in Computer and Information Sciences, Teaching assistant at Computer Science Department Faculty of Computer and Information Sciences Ain Shams University Supervised By Prof. Dr. El-Sayed M. El-Horbaty Dr. Abeer M. Mahmoud Cairo 2015
  • 2. AGENDA  Introduction  Problem Definition  Objectives  Related Work  Methodologies  Conclusions & Future Work  Publications  References
  • 3. AGENDA  Introduction  Problem Definition  Objectives  Related Work  Methodologies  Conclusions & Future Work  Publications  References
  • 5. GRAYSCALE IMAGE  Monochrome / One color image  No color info  Represent the brightness of the image  8 bits/pixel data  256 different brightness level
  • 6. MEDICAL IMAGE  Used to create images of the human body  Seeks to  Reveal internal structures hidden by the skin and bones  Diagnose and treat disease PET CT MRI
  • 7. DEGRADATION MODEL  Noises  Domains  Filters 𝒈(𝒙, 𝒚) 𝒇^ (𝒙, 𝒚) Noise 𝜼(𝒙, 𝒚) + DEGRADATION RESTORATION 𝒇(𝒙,𝒚) Degradation function H Restoration Filter(s)
  • 8. GAUSSIAN NOISE  Additive  𝑝 𝑧 = 1 2𝜋 𝛿 𝑒−(𝑧−𝑧`) 2𝛿2  Due to:  Poor illumination  High temperature
  • 9. SALT & PEPPER NOISE  Additive  𝑝 𝑧 = 𝑃𝑎 𝑓𝑜𝑟 𝑧 = 𝑎 𝑃𝑏 𝑓𝑜𝑟 𝑧 = 𝑏 0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒  Due To:  Quick transients  Data transmission errors
  • 10. IMAGE ENHANCEMENT & RESTORATION  Attempts to recover an image that has been distorted by using a priori knowledge of the degradation process. 𝒈(𝒙, 𝒚) 𝒇^ (𝒙, 𝒚) Noise 𝜼(𝒙, 𝒚) + DEGRADATION RESTORATION 𝒇(𝒙,𝒚) Degradation function H Restoration Filter(s)
  • 11. ENHANCEMENT DOMAINS  Spatial Domain  Transform Domain (Frequency Domain)
  • 12. SPATIAL DOMAIN  𝑔(𝑥, 𝑦) = 𝑇[𝑓(𝑥, 𝑦)]  Direct manipulation of pixels in an image  Covers the neighborhood operations (Convolution)  Filters in this domain are time consuming
  • 13. SPATIAL FILTERING Filters Linear Arithmetic Geometric Harmonic Contra- Harmonic Non-Linear Order Statistics Min Max Midpoint Median Alpha trimmed Adaptive Local Noise Reduction Adaptive Median
  • 14. TRANSFORM DOMAIN (FREQUENCY DOMAIN)  𝐹(𝑢, 𝑣) = 𝑅[𝑓(𝑥, 𝑦)]  𝑔(𝑥, 𝑦) = 𝑅−1 𝑇[𝐹(𝑢, 𝑣)]  A representation of an image as a sum of complex exponentials of varying magnitudes, frequencies, and phases
  • 15. FREQUENCY FILTERING  Filters in spatial domain have their corresponding form in frequency domain  Convolution in the spatial domain corresponds to multiplication in the frequency domain, and vice versa  Wiener filter  Gaussian removal  De-blurring
  • 16. QUALITY METRICS  Used for dynamically monitor and adjust image quality  MSE = 1 𝑀𝑁 𝑖 𝑀 𝑗 𝑁 [𝑔 𝑖, 𝑗 − 𝑓(𝑖, 𝑗)]2  𝑆𝑁𝑅 = 10 𝑙𝑜𝑔10 ( 𝑖 𝑀 𝑗 𝑁 𝑓(𝑖,𝑗) 𝑖 𝑀 𝑗 𝑁[𝑔 𝑖,𝑗 −𝑓(𝑖,𝑗)]2)  𝑃𝑆𝑁𝑅 = 10 𝑙𝑜𝑔10 ( 2552 𝑀𝑆𝐸 )
  • 18. PARALLEL ALGORITHM FEATURES  Granularity  Coarse-grained  Fine-grained  Type of parallel processing  Explicit  Implicit  Synchronization  Latency  Scalability
  • 19. DECOMPOSITION CLASSIFICATIONS  Data Decomposition – (Fine Grained) Data D1 D2 D3 D4 D5 … Dn
  • 20. DECOMPOSITION CLASSIFICATIONS  Task (Functional) Decomposition – (Coarse Grained) Code Block (Group of Functions) Fn1 Fn2 Fn3 Fn4 Fn5 … Fnn
  • 22. METRICS  Speed Up  𝑆 𝑃 = 𝑇𝑠 𝑇𝑝  Efficiency  𝐸𝐹𝐹 = 𝑆 𝑃 𝑃
  • 23. AGENDA  Introduction  Problem Definition  Objectives  Related Work  Methodologies  Conclusions & Future Work  Publications  References
  • 24. PROBLEM DEFINITION  Images are corrupted by multi-noise type (i.e. Gaussian, salt & pepper) at the same time.  There is an open demand for multi-noise removal filters.
  • 25. PROBLEM DEFINITION  Medical images have very large size  Processing these forms take so much time to process sequentially  We have to parallelize the traditional sequential algorithms for the sake of time performance improvement.
  • 26. AGENDA  Introduction  Problem Definition  Objectives  Related Work  Methodologies  Conclusions & Future Work  Publications  References
  • 27. OBJECTIVES 1. Study the literature work and implement Gaussian de-noising experiment as a case study.
  • 28. OBJECTIVES 2. Design & Develop sequential hybrid de- noising filter for multi- noise removal and compare between it and the simple filters in terms of peak signal to noise ratio (PSNR).
  • 29. OBJECTIVES 3. Design & Develop a parallel hybrid de-noising filter then compare it relative to the sequential hybrid above in terms of time.
  • 30. AGENDA  Introduction  Problem Definition  Objectives  Related Work  Methodologies  Conclusions & Future Work  Publications  References
  • 31. RELATED WORK  Single Noise Removal  Multi-Noise Removal  Parallel Image Processing
  • 33. Year Author Contribution Noise to Remove Input Quality Criteria 2014 Monika P. and Sukhdev S Comparative study on images de-noising techniques Salt & pepper Lena Image MSE PSNR 2014 Rupinder Kaur, Prabhpreet Kaur Novel technique and comparison Speckle Ultrasound image SNR PSNR CoC EPI 2014 P.Deepa and M.Suganthi Overview on image de- noising techniques Gaussian Lung CT image MSE PSNR 2014 Sezal Khera and Sheenam Malhotra Comparative study for de-nosing techniques Gaussian, Slat & Pepper, Poison and Speckle MRI, X-Ray, CT PSNR
  • 35. Year Authors Distortion Hybrid Input Quality Criteria 2014 Seema and Meenakshi G. Speckle + Gaussian Median Filter based on DWT via soft thresholding + Mean absolute difference Standard gray scale images PSNR 2013 Ankita D. and et.al Camera & object motion blurs + Gaussian + Salt & pepper Wiener + Median Gradient images PSNR MSE 2013 Versha R. and Priyanka K. Gaussian + impulsive + Speckle + Possion Curvelet transform + Unsharp Mask filter + Median filter One gray & One colored images PSNR 2012 J U. and G R. Gaussian + Impulsive Haar wavelet filter + soft thresholding technique + center weighted median 10 DICOM images PSNR MAE UQI ET
  • 37. Year Author(s) Contribution Techniques 2015 Sharanjit Singh and et.al Implementation 1. Functions Extractions of Images 2. Clustering of 1 3. Histogram generation of 2 2014 Er.Paramjeet kaur and Er.Nishi Analysis CUDA, DicrectCompute, OpenCL 2013 Sanjay Saxena and et.al Implementation Algorithms 1. Image Segmentation 2. Image de-noising 3. Histogram Equalization
  • 39. GAUSSIAN DE-NOISING ALGORITHMS Filters Linear Arithmetic Geometric Harmonic Contra- Harmonic Non-Linear Order Statistics Min Max Midpoint Median Alpha trimmed Adaptive Local Noise Reduction Adaptive Median
  • 40. ENVIRONMENT SETTINGS  Input:  Mode: Gray scale – no color info  Dimensions: average 512x512  Distortion  Gaussian Noise  Mean = 0 &Variance = 1000  Filter Size  3x3, 5x5, 7X7 and 9x9  Machine Specs:  Processor: Intel core i5  RAM: 4 GB  OS: Windows 7 home edition, 64 bit  Language: C#  Tool:
  • 41. SAMPLE RUN Geometric Harmonic Midpoint Alpha Trimmed Local Noise Reduction GroundTruth Of 3X3
  • 46. AGENDA  Introduction  Problem Definition  Objectives  Related Work  Methodologies  Conclusions & Future Work  Publications  References
  • 47. METHODOLOGIES 1. Serial Hybrid Algorithm 2. Parallel Hybrid Algorithm
  • 48. SERIALHYBRIDALGORITHM Fourier Transform Wiener Inverse Fourier Transform Adaptive Median Add Gaussian Add Salt & Pepper Circular Blur Image Noisy Image Power Spectrum Restored Image Spatial Domain Fourier Transform Domain
  • 49. PSEUDO-CODE original = read image from file original = convert to double values noisyBuf = degradImage(original) procedure hybridFilter (noisyBuf) adapOnly = apply adaptive median filter adapMedian(noisyBuf, 7 pix) hybrid = apply wiener filter wiener(adapOnly, original) hybrid = apply contrast(hybrid) end procedure procedure degradeImage(original) I1 = Add circular blur to original I2 = Add Gaussian noise to I1 I3 = Add Salt & Pepper noise to I3 return I3 as Noisy buffer end procedure 1
  • 50. PSEUDO-CODE original = read image from file original = convert to double values noisyBuf = degradImage(original) procedure hybridFilter (noisyBuf) adapOnly = apply adaptive median filter adapMedian(noisyBuf, 7 pix) hybrid = apply wiener filter wiener(adapOnly, original) hybrid = apply contrast(hybrid) end procedure procedure adapMedian(g, Smax) f = g; %initial setup alreadyProcessed = false(size(g)); for all window size k = start from 3 to Smax do: get Min, max and median Values of window and store it in zmin, zmax and zmed; determine if we will use level b processing in processByLvlB; zB = (g > zmin) & (zmax > g); outputZxy = processByLvlB & zB; outputZmed = processByLvlB & ~zB; f(outputZxy) = g(outputZxy); f(outputZmed) = zmed(outputZmed); alreadyProcessed = alreadyProcessed or processUsingLevelB; if all alreadyProcessed array has been done break the loop end if end for fill the remaining values in alreadyProcessed if exist by the zmed end procedure 1 2
  • 51. PSEUDO-CODE original = read image from file original = convert to double values noisyBuf = degradImage(original) procedure hybridFilter (noisyBuf) adapOnly = apply adaptive median filter adapMedian(noisyBuf, 7 pix) hybrid = apply wiener filter wiener(adapOnly, original) hybrid = apply contrast(hybrid) end procedure procedure wiener(adapOnly, original) F = create power spectrum of the adapOnly sigma_u = calculate noise sigma in F output = wiener deconvolution for F given the original image and sigma_u end procedure 1 2 3
  • 52. PSEUDO-CODE original = read image from file original = convert to double values noisyBuf = degradImage(original) procedure hybridFilter (noisyBuf) adapOnly = apply adaptive median filter adapMedian(noisyBuf, 7 pix) hybrid = apply wiener filter wiener(adapOnly, original) hybrid = apply contrast(hybrid) end procedure procedure contrast(image) stretch image’s pixels values to fit in range 0 - 255 end procedure 1 2 3 4
  • 53. ENVIRONMENT SETTINGS  Input:  Mode: Gray scale – no color info  Dimensions:  512 X 512 (0.5 MB)  128 X 128 (17 KB)  Resolution: 300px  Distortion  Gaussian  Salt & Pepper  Circular Blur  Machine Specs:  Processor: Intel core i5  RAM: 4 GB  OS: Windows 7 home edition, 64 bit  Tool:
  • 54. RESULTS  Average PSNR for Adaptive, Wiener and Serial Hybrid Image a b c d e f g h I J k l AVG Adaptive 10.04 14.15 14.2 12.06 13.3 13.7 13.43 9.5 13.29 17.12 6.88 12.45 12.6 Wiener 11.63 23.37 18.63 14.37 16.46 8.79 12.39 11.6 15.06 22.53 11.92 16.73 15.3 Hybrid 13.36 29.7 26.7 22.75 19.32 18.34 17.66 18.84 18.97 26.04 24.69 20.64 19.8
  • 55. 0 5 10 15 20 25 30 35 a b c d e f g h i j k l PSNR IMAGE PSNR chart for full size images for proposed hybrid approach, adaptive & wiener Hybrid Wiener Adaptive
  • 56. 0 5 10 15 20 25 30 35 a b c d e f g h i j k l PSNR IMAGE Thumbnails VS. full Sizes of PSNR values Thumbnails Full Sizes
  • 57. 0 10 20 30 40 50 60 a b c d e f g h i j k l TIME(MINS) IMAGE Thumbnails VS. full Sizes time plot in minutes Thumbnails Full Size
  • 58. PARALLEL HYBRID ALGORITHM  Adaptive Median takes too much time  Spatial domain  Window size increases Fourier Transform Wiener Inverse Fourier Transform Adaptive Median Add Gaussian Add Salt & Pepper Circular Blur Image Noisy Image Power Spectrum Restored Image Spatial Domain Fourier Transform Domain
  • 59. PARALLELHYBRIDALGORITHM Fourier Transform Wiener Inverse Fourier Transform Adaptive Median Add Gaussian Add Salt & Pepper Circular Blur Image Noisy Image Power Spectru m Restored Image Spatial Domain Fourier Transform Domain
  • 60. PARALLELHYBRIDALGORITHM Fourier Transform Wiener Inverse Fourier Transform Adaptive Median Add Gaussian Add Salt & Pepper Circular Blur Image Noisy Image Power Spectru m Restored Image Spatial Domain Fourier Transform Domain
  • 61. PARALLELHYBRIDALGORITHM Fourier Transform Wiener Inverse Fourier Transform Adaptive Median Add Gaussian Add Salt & Pepper Circular Blur Image Noisy Image Power Spectru m Restored Image Spatial Domain Fourier Transform Domain
  • 62. PSEUDO-CODE original = read image from file original = convert to double values noisyBuf = degradImage(original) procedure parallelHybridFilter (noisyBuf) adapOnly = apply adaptive median filter parAdapMedian(noisyBuf, 11 pix, workersNo) hybrid = apply wiener filter wiener(adapOnly, original) hybrid = apply contrast(hybrid) end procedure procedure degradeImage(original) I1 = Add circular blur to original I2 = Add Gaussian noise to I1 I3 = Add Salt & Pepper noise to I3 return I3 as Noisy buffer end procedure 1
  • 63. PSEUDO-CODE original = read image from file original = convert to double values noisyBuf = degradImage(original) procedure parallelHybridFilter (noisyBuf) adapOnly = apply adaptive median filter parAdapMedian(noisyBuf, 11 pix, workersNo) hybrid = apply wiener filter wiener(adapOnly, original) hybrid = apply contrast(hybrid) end procedure procedure parAdapMedian(g, Smax, workersNo) for each worker in workersNo do in Parallel apply adaptive median filter convolution procedure on g end for end procedure 1 2
  • 64. PSEUDO-CODE original = read image from file original = convert to double values noisyBuf = degradImage(original) procedure parallelHybridFilter (noisyBuf) adapOnly = apply adaptive median filter parAdapMedian(noisyBuf, 11 pix, workersNo) hybrid = apply wiener filter wiener(adapOnly, original) hybrid = apply contrast(hybrid) end procedure procedure wiener(adapOnly, original) F = create power spectrum of the adapOnly sigma_u = calculate noise sigma in F output = wiener deconvolution for F given the original image and sigma_u end procedure 1 2 3
  • 65. PSEUDO-CODE original = read image from file original = convert to double values noisyBuf = degradImage(original) procedure parallelHybridFilter (noisyBuf) adapOnly = apply adaptive median filter parAdapMedian(noisyBuf, 11 pix, workersNo) hybrid = apply wiener filter wiener(adapOnly, original) hybrid = apply contrast(hybrid) end procedure procedure contrast(image) stretch image’s pixels values to fit in range 0 - 255 end procedure 1 2 3 4
  • 66. ENVIRONMENT SETTINGS  Input:  Mode: Gray scale – no color info - 500px  Dimensions:  1900 x 2368 (277 KB)  3800 x 4736 (715 KB)  15200 x 18944 (2.5 MB)  Distortion:  Gaussian  Salt & Pepper  Circular Blur  Machine Specs:  Processor: Intel core i5  RAM: 6 GB  OS: Windows 7 enterprise edition, 64 bit  Tool
  • 67. RESULTS  Time in seconds for the 3 image sizes each of which divided into 2,4 an 8 partitions  Workers rang 2 - 12 Image Dimensions Serial Workers 2 4 6 8 10 12 1900 X 2368 79.5 2-Partitions 10.41 10.87 10.89 10.37 10.96 11.11 4- Partitions 12.1 11.99 12.22 12.29 12.49 12.23 8- Partitions 16.22 15.84 15.87 16.62 15.64 16.2 3800 X 4736 272.42 2- Partitions 44.67 41.59 36.14 35.12 42.86 49.28 4- Partitions 41.37 40.76 40.62 41.06 41.35 42.18 8- Partitions 54.88 54.65 56.63 56.85 54.69 55.21 7600 X 9472 1742.5 2- Partitions 640.34 863.87 856.24 652.75 914.24 690.8 4- Partitions 768 470.08 480.03 532.75 490.92 703.38 8- Partitions 847.9 538.55 769.91 505.67 656.4 611.59
  • 68. 0 100 200 300 400 500 600 700 800 2 4 6 8 10 12 TIME(SECONDS) WORKER Average time consumed in seconds for serial and parallel 2,4 and 8 partition input Serial 2-Partition 4-Partition 8-Partition
  • 69. 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 2 4 6 8 10 12 SPEEDUP WORKER Average speed up for 2, 4 and 8 partition input SP -2 Sp - 4 sp - 8
  • 70. 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 2 4 6 8 10 12 EFFICIENCY WORKER Efficiency for 2, 4 and 8 partition input EFF -2 EFF - 4 EFF - 8
  • 71. AGENDA  Introduction  Problem Definition  Objectives  Related Work  Methodologies  Conclusions & Future Work  Publications  References
  • 72. CONCLUSIONS  Simple Gaussian filters comparison showed for the average run 3, 5, 7 and 9 pix that  Harmonic filter gave the max. PSNR  Geometric filter gave the min. Time
  • 73. CONCLUSIONS  Presented serial hybrid filter gave PSNR enhancement  33.33% VS. Wiener  55.55% VS. Adaptive median
  • 74. CONCLUSIONS  Parallel hybrid filter gave a speed up  3x for 2-partition  4x for 4-partition  3.5x for 8-partition
  • 75. FUTURE WORK  Apply the proposed methods medical imaging modalities like PET (Colored).  Try to corrupt the image with higher noise probabilities.  Change the noise type to but take care that it suits the selected adaptive median and wiener filters. IMAGE PROCESSING
  • 76. FUTURE WORK  Try Different parallel model  GPU  CPU + GPU (Hybrid Parallel).  Compare time performance with the existing methods.  Integrate with cloud computing. PARALLEL PROCESSING
  • 77. AGENDA  Introduction  Problem Definition  Objectives  Related Work  Methodologies  Conclusions & Future Work  Publications  References
  • 78. PUBLICATIONS  Nora Youssef, Abeer M.Mahmoud and EL- Sayed M.EL-Horbaty “Gaussian De- Noising Techniques in Spatial Domain for Gray Scale Medical Images”, Int. J. of Information Technologies and Knowledge (ITK), Vol. 8, No.3, pp.90-100, Bulgaria, Jun. 2014.
  • 79. PUBLICATIONS  Nora Youssef, Abeer M.Mahmoud and EL-Sayed M.EL-Horbaty “A Hybrid De- Noising Technique for Multi-noise Removal on Gray Scale Medical Images”, Int. J. of Tomography and Simulation (IJTS) IF 0.75, Vol. 28, No.2, pp 106-116, India, Mar. 2015.
  • 80. PUBLICATIONS  Nora Youssef, Abeer M.Mahmoud and EL- Sayed M.EL-Horbaty “Gaussian De-Noising Techniques” Lambert Academic Publishing, Germany, Apr. 2015.
  • 81. PUBLICATIONS  Nora Youssef, Abeer M.Mahmoud and EL-Sayed M.EL-Horbaty “A Parallel Hybrid Technique for Multi-Noise Removal from Grayscale Medical Images” Int. J Real Time Image Processing, Springer, Berlin, May 2015 (Submitted).
  • 82. AGENDA  Introduction  Problem Definition  Objectives  Related Work  Methodologies  Conclusions & Future Work  Publications  References
  • 83. REFERENCES  Ravi M. Rai and Urooz J. “Analysis Techniques For Eliminating Noise In Medical Images Using Bivariate Shrinkage Method” Int. J. of Advanced Research in Computer Engineering & Technology vol. 2, no. 10, pp. 2737 to 2740, 2013.  Ankita D. and Archana S. “An Advanced Filter for Image Enhancement and Restoration” J. Open Journal of Advanced Engineering Techniques OJAET vol.1, no. 1, pp. 7-10, 2013.  Salem Al-amri, N. V. Kalyankar and S. D. Khamitkar “A Comparative Study of Removal Noise from Remote Sensing Image” Int. J. of Computer Science Issues, vol. 7, no. 1, pp. 32 - 35, 2010.  Sanjay S., Neeraj S. and Shiru S. "Image Processing Tasks using Parallel Computing in Multi core Architecture and its Applications in Medical Imaging" Int. J. of Advanced Research in Computer and Communication Engineering vol. 2, no. 4, pp. 1896 to 1899, 2013
  • 84. REFERENCES  Rajeshwari S., Sharmila T. Sree “Efficient quality analysis of MRI image using preprocessing techniques,” Proceedings of 2013 IEEE Conference on Information and Communication Technologies (ICT), pp.: 391-396, 2013.  Rafael C. Gonzalez and Richard E. Woods “Digital Image Processing” Person Education,3rd Edition, 2008.  Nikola R. and Milan T. “Improved Adaptive Median Filter for Denoising Ultrasound Images” Conf. Advances in Computer Science pp. 196 - 174, 2012  P.Deepa and M.Suganthi “Performance Evaluation of Various Denoising Filters for Medical Image”, IJCSIT, Vol. 5, No.3, pp. 4205-4209 , 2014
  • 85. REFERENCES  B.Mohd. Jabarullah, Sandeep Saxena and Dr.C. Nelson Kennedy Badu“Survey on Noise Removal in Digital Images” Vo. 6, No. 4, pp.45 -51 2012  Gajanand G.“Algorithm for Image Processing Using Improved Median Filter and Comparison of Mean, Median and Improved Median Filter” IJSCE, vol. 1, no.5 pp. 304 – 311, 2011.  K.Selvanayaki, Dr. M. Karnan “CAD System for Automatic Detection of Brain Tumor through Magnetic Resonance Image-A Review” Int. J. of Engineering Science and Technology vol. 2, no.10, pp. 5890-5901, 2010.  Monika P. and Sukhdev S. “Comparative Analysis of Image Denoising Techniques” Int. J. of Computer Science & Engineering Technology (IJCSET) vol. 5 no. 02 pp. 160 - 167, 2014
  • 88. APPENDICES  Serial Code  Parallel Code  Data Sets  Publications Materials
  • 89. original = imread(imPath); original = im2double(original); noisyBuf = degradImage(original); procedure hybrid (noisyBuf) adapOnly = apply adaptive median filter (noisyBuf, 7); hybrid = apply wiener filter (adapOnly, original); hybrid = apply contrast(hybrid); end procedure procedure adapMedian(g, Smax) f = g; %initial setup alreadyProcessed = false(size(g)); for all window size k = start from 3 to Smax do: zmin = get Min Value of window; zmax = get Max Value of window; zmed = get Median Value of window; processByLvlB determine if we will use level b processing; zB = (g > zmin) & (zmax > g); outputZxy = processByLvlB & zB; outputZmed = processByLvlB & ~zB; f(outputZxy) = g(outputZxy); f(outputZmed) = zmed(outputZmed); alreadyProcessed = alreadyProcessed or processUsingLevelB; if all alreadyProcessed array has been done break the loop end if end for fill the remaining values in alreadyProcessed if exist by the zmed end procedure SERIAL CODE
  • 90. PARALLEL CODE original = imread(imPath); original = im2double(original); noisyBuf = degradImage(original); procedure parallelHybrid (noisyBuf) adapOnly = parallelAdapMedian(noisyBuf, 11, partition_i); partition_i belongs to {2, 4, 8}; hybrid = apply Wiener function(adapOnly, original); hybrid = apply contrast(hybrid); end procedure procedure parallelAdapMedian(g, Smax, parts) [M, N] = size(g); f = g; %initial setup alreadyProcessed = false(size(g)); splittedImg = imgSpliter( f, M, N, parts ); for subimg = 1 into parts, numberOfWorkers_i do in parallel numberOfWorkers_i belongs to {2, 4, 6 , 8, 10, 12} myTemp = splittedImg(subimg); t = alreadyProcessed; for k = 3 to Smax do zmin = get Min. value in subimg zmax = get Max. value in subimg zmed = get Median value in subimg processByLvlB = determine if we will use level b processing zB = (g > zmin) & (zmax > g); outputZxy = processByLvlB & zB; outputZmed = processByLvlB & ~zB; myTemp(outputZxy) = g(outputZxy); myTemp(outputZmed) = zmed(outputZmed); end for splittedImg(subimg) = myTemp; end parfor end procedure
  • 91. DATA SETS  idoimagaing  http://idoimaging.com/wiki/tiki- index.php?page=Sample+Data  3DISC  http://www.3discimaging.com/our-products/medical- solutions/firecr-medicalreaders/image-quality/  Computed Tomography Emphysema Database  http://image.diku.dk/emphysema_database/  Some of images are available on Google Images.
  • 93. GAUSSIAN DE-NOSING TECHNIQUES IN SPATIAL DOMAIN FOR GRAY SCALE MEDICAL IMAGES Published International Journal "Information Technologies & Knowledge" Volume 8  Number 3 ISSN 1313-0455 Bulgaria 2014 Click here…
  • 94. A HYBRID DE-NOISING TECHNIQUE FOR MULTI-NOISE REMOVAL ON GRAY SCALE MEDICAL IMAGES Published International Journal of Tomography and Simulation Volume, 28  Number, 2 ISSN, 2319-3336 IF 0.75 India 2015 Click Here…
  • 95. GAUSSIAN DE-NOSING TECHNIQUES Published Book Chapter ISBN: 978-3-659-49570-0 Lambert Academic Publishing (LAP) Germany 2015 Click Here…
  • 96. A PARALLEL HYBRID TECHNIQUE FOR MULTI-NOISE REMOVAL FROM GRAYSCALE MEDICAL IMAGES Submitted Int. J. Real Time Image Processing Springer IF: 1.1 Germany 2015 Click Here ...

Editor's Notes

  1. PET - Positron Emission Tomography CT - Computerized Tomography MRI - Magnetic Resonance Imaging Ref: MNT Knowledge Center web link http://www.medicalnewstoday.com/articles/146309.php , Access date 9/1/2014
  2. Green are commonly used for Gaussian removal Red are commonly used for Sal & Pepper removal
  3. Figure: Original image vs. Fourier magnitude representation
  4. MSE  Mean Square Error SNR  Signal to Noise Ratio PSNR  Peak Signal to Noise Ratio
  5. Coarse-grained  Few tasks but have large computations (Functional Decomposition) Fine-grained  Many tasks but have small computations (Data Decomposition)
  6. MSE  Mean Square Error PSNR  Peak Signal to Noise Ratio CoC  Correlation Coefficients EPI  Edge Preservation Index MAE  Mean Absolute Error UQI  Universal Quality Index ET  Execution Time (It was abbreviated by the paper’s Authors) ---- Ref to Book Chapter page 45 for full index
  7. MSE  Mean Square Error PSNR  Peak Signal to Noise Ratio CoC  Correlation Coefficients EPI  Edge Preservation Index MAE  Mean Absolute Error UQI  Universal Quality Index ET  Execution Time (It was abbreviated by the paper’s Authors) ---- Ref to Book Chapter page 45 for full index
  8. Sample run of 3x3 pix filter size A b C d E from paper
  9. Correctness 1. Initially we begin a 2D image matrix (I) with size NxN distorted by 3 types of noise, Blur (B), Gaussian (G) and Salt & Pepper (S&P). 2. The Selected adaptive median (am) filter will scan (I) pixel by pixel in a sequential manner for S&P removal. 3. (am) filter will terminate after a finite iterations based on (I) size NxN and subwindow size mxm. 4. The selected Wiener (w) filter will scan the image power spectrum (PS) for (B) and (G) removal and terminates on the end of the PS. 5. The whole filter will terminate after (w) filter termination as it is the last step in the hybrid filter.
  10. Adaptive Median is the bottleneck Because it runs on spatial domain and the filter on the worst case runs 3, 5, 7, 9, 11 px so that the filter size increases.
  11. Adaptive Median is the bottleneck Because it runs on spatial domain and the filter on the worst case runs 3, 5, 7, 9, 11 px so that the filter size increases.
  12. Adaptive Median is the bottleneck Because it runs on spatial domain and the filter on the worst case runs 3, 5, 7, 9, 11 px so that the filter size increases.
  13. Adaptive Median is the bottleneck Because it runs on spatial domain and the filter on the worst case runs 3, 5, 7, 9, 11 px so that the filter size increases.
  14. Correctness 1. Initially we begin a 2D image matrix (I) with size NxN distorted by 3 types of noise, Blur (B), Gaussian (G) and Salt & Pepper (S&P). 2. Input image will be divided into k partitions. 3. The Selected adaptive median (am) filter remove the S&P noise from the k partitions in parallel. 4. (am) filter will terminate after a finite iterations based on I size NxN and subwindow size mxm. 5. The selected Wiener (w) filter will scan the image power spectrum (PS) for B and G removal and terminates on the end of the PS. 6. The whole filter will terminate scan the k partitions of the image and (w) filter termination.
  15. Speed Up 𝑆 𝑃 = 𝑇 𝑠 𝑇 𝑝
  16. Efficiency 𝐸𝐹𝐹= 𝑆 𝑃 𝑃
  17. Link to ver0.4 after Dr.Sayed revision