Useful Info

International Journal of Innovative Research in Advanced Engineering (IJIRAE) ISSN: 2349-2163
Volume 1 Issue 7 (August 2014 ) http://ijirae.com
An FPGA Based Image Denoising Architecture with Histogram

Equalization for Enhancement
Jayalakshmi S Nair* Bibin Binu Simon
ECE Dept. & M.G. University ECE Dept. &M.G.University
Abstract— This work provides an insight into the design of an efficient denoising architecture for removal of impulse noise in
images. Images are often corrupted by impulse noise in the procedures of image acquisition and transmission. In this work an
efficient denoising scheme and its VLSI architecture for the removal of random-valued impulse noise is proposed. To achieve
the goal of low cost, a low complexity VLSI architecture is proposed. Here a decision-tree-based impulse noise detector to
detect the noisy pixels, and an edge-preserving filter to reconstruct the intensity values of noisy pixels are used. Furthermore,
histogram equalization is used to enhance the effects of removal of impulse noise. Our extensive experimental results
demonstrate that the proposed technique can obtain better performances in terms of both quantitative evaluation and visual
quality than the previous lower complexity methods. Moreover, the performance can be comparable to the higher complexity
methods. Its hardware cost is low and suitable to be applied to many real-time applications. These are implemented using
Verilog Hardware Description Language. The Verilog code is synthesized on Spartan 3 FPGA using Xilinx ISE 14.7 software
tool.
Keywords— FPGA, Image denoising, impulse noise, impulse detector, Xilinx

I. INTRODUCTION
Image processing is widely used in many fields, such as medical imaging, scanning techniques, printing skills, license plate
recognition, face recognition, and so on. In general, images are often corrupted by impulse noise in the procedures of image
acquisition and transmission. The noise may seriously affect the performance of image processing techniques. Hence, an efficient
denoising technique becomes a very important issue in image processing. According to the distribution of noisy pixel values,
impulse noise can be classified into two categories: fixed valued impulse noise and random-valued impulse noise. The former is
also known as salt-and-pepper noise because the pixel value of a noisy pixel is either minimum or maximum value in gray-scale
images. The values of noisy pixels corrupted by random-valued impulse noise are uniformly distributed in the range of [0, 255] for
gray-scale images. There have been many methods for removing salt and-pepper noise, and some of them perform very well. The
random-valued impulse noise is more difficult to handle due to the random distribution of noisy pixel values. Here we only focus
on removing the random-valued impulse noise from the corrupted image in this paper.
Recently, many image denoising methods have been proposed to carry out impulse noise suppression. Some of them employ
the standard median filter or its modifications. However, these approaches might blur the image since both noisy and noise-free
pixels are modified. To avoid the damage on noise-free pixels, an efficient switching strategy has been proposed in the literature.
In general, the switching median filter consists of two steps: 1) impulse detection and 2) noise filtering. It locates the noisy pixels
with an impulse detector, and then filters them rather than the whole pixels of an image to avoid causing the damage on noise-free
pixels. In addition to median filter, there are other methods used to carry out impulse noise.
Generally, the denoising methods can be classified into two categories: lower complexity techniques and higher complexity
techniques. The complexity of denoising algorithms depends mainly on the local window size, memory buffer, and iteration times.
The lower complexity techniques use a fixed-size local window, require a few line buffers, and perform no iterations. Therefore,
the computational complexity is low. However, the reconstructed image quality is not good enough. The higher complexity
techniques yield visually pleasing images by using high computational complexity arithmetic operations, enlarging local window
size adaptively or doing iterations. The higher complexity approaches require long computational time as well as full frame buffer.
Digital image are often corrupted by different types of noise, namely, additive white Gaussian noise, impulse noise and mixed
(Gaussian and impulse) noise . Noises are added in the image during acquisition by camera sensors and transmission in the
channel. Hence, image de-noising is one of the most common and important image processing operations in image and video
processing applications. Today, in many practical real-time applications, the denoising process is included in end-user equipment,
so there appears an increasing need of a good lower-complexity denoising technique, which is simple and suitable for low-cost
VLSI implementation. Low cost is a very important consideration in purchasing consumer electronic products. To achieve the goal
of low cost, less memory and easier computations are indispensable. In this paper, we focus only on the lower complexity
denoising techniques because of its simplicity and easy implementation with the VLSI circuit.
Although the field of digital image processing is built on a foundation of mathematical and probabilistic formulation human
eyes and analysis play a central role in the choice of one technique versus another, and this choice is often made based on
subjective and visual judgments. The impulse noise called salt and pepper causes black and white points appears in digital gray
scale images, which chaotically scattered along image area
In order to reduce noise in digital images, a great number of algorithms have been presented. One of the most common
ways to remove noise is convolving the image with a mask that represents a low-pass filter (e.g. Gaussian filter), performing a
smoothing operation through a weighted average of neighbours, where the weights decrease with distance from the central pixel
________________________________________________________________________________________________________
© 2014, IJIRAE- All Rights Reserved Page - 255
on filter window. Often, the use of smoothing filters cause significant blurring of edges. Recently, Buades et al. proposed
interesting work in which they uses the Non-Local Means algorithm, supported on the idea that images contain repeated structures,
and averaging this structures, noise can be reduced. The decision tree is a simple but powerful form of multiple variable analysis .
It can break down a complex decision-making process into a collection of simpler decisions, thus provide a solution which is often
easier to interpret. There have been several methods using decision tree to deal with salt-and-pepper noise and some of them
perform well. Based on above basic concepts, we present a novel adaptive decision-tree-based denoising method (DTBDM) and
its VLSI architecture for removing random-valued impulse noise. To enhance the effects of removal of impulse noise, the results
of reconstructed pixels are adaptively written back as a part of input data.
The proposed design requires simple computations and two line memory buffers only, so its hardware cost is low. For a
512 x 512 x 8-bit gray scale test image, only two line buffers (512 x 2 x 8 bits) is needed in our design. Most state-of-the-art
methods need to buffer a full image (512 x 512 x 8) bits. In the proposed design, 99.6 percent of storage is reduced. Furthermore,
only simple arithmetic operations, such as addition and subtraction, are used in DTBDM. Especially, it can remove the noise from
corrupted images efficiently and requires no previous training. These experimental results demonstrate that the proposed technique
can obtain better performances in terms of both quantitative evaluation and visual quality than other lower complexity denoising
methods.
Moreover, the performance can be comparable to the higher complexity methods. The seven-stage VLSI architecture for
the proposed design will be implemented and synthesized by using Verilog HDL and Xilinx ISE Tool using Spartan 3 FPGA
device respectively. The content of this report is organized as follows. First a literature survey on existing methods is described
briefly. The proposed DTBDM is introduced briefly in next section and describes the proposed VLSI architecture in detail. Next
section illustrates the VLSI implementation and comparisons. The conclusion is provided in next section with advantages and
disadvantages along with scope of the proposed work.
II. DECISION TREE BASED DENOISING METHOD

Many denoising schemes are “decision-based” median filters. This means that the noise candidates are first detected by
some rules and are replaced by the median output or its variants. These schemes are good because the uncorrupted pixels will not
be modified. However, the replacement methods in these denoising schemes cannot preserve the features of the images, in
particular the edges are smeared. a decision based algorithm is proposed. In this, image is denoised by using a 3×3 window. If the
processing pixel value is 0 or 255 it is processed or else it is left unchanged. The selected 3× 3 window elements are arranged in
either increasing or decreasing order. Then the pixel values 0’s and 255’s in the image (i.e., the pixel values responsible for the salt
and pepper noise) are removed from the image. Then the median value of the remaining pixels is taken. This median value is used
to replace the noisy pixel.
A .Decision-Tree-Based Impulse Detector

The proposed method here is decision tree based denoising method (DTBDM). The noise considered in this paper is random-
valued impulse noise with uniform distribution . Here, we adopt a 3 x 3 mask for image denoising. Assume the pixel to be
denoised is located at coordinate (i, j) and denoted as pi,j, and its luminance value is named as f i,j, as shown in Figure.1. According
to the input sequence of image denoising process, we can divide other eight pixel values into two sets: WTopHalf and WBottomHalf .
They are given as
WTopHalf = {a, b, c, d} (1)
WBottomHalf = { e, f, g, h (2)
Fig. 1 A 3X 3mask centered on pi,j
DTBDM consists of two components: decision-tree-based impulse detector and edge-preserving image filter. The
detector determines whether p i,j is a noisy pixel by using the decision tree and the correlation between pixel p i,j and its
neighbouring pixels. If the result is positive, edge preserving image filter based on direction-oriented filter generates the
reconstructed value. Otherwise, the value will be kept unchanged. The design concept of the DTBDM is displayed in Figure.2.
In order to determine whether pi,j is a noisy pixel, the correlations between and its pi,j neighbouring pixels are considered .
Surveying these methods, we can simply classify them into several ways—observing the degree of isolation at current pixel
determining whether the current pixel is on a fringe or comparing the similarity between current pixel and its neighbouring pixels.
Therefore,in the decision-tree-based impulse detector, we design three modules—isolation module (IM), fringe module (FM), and
similarity module (SM). Three concatenating decisions of these modules build a decision tree. The decision tree is a binary tree
and can determine the status of pi,j by using the different equations in different modules.
________________________________________________________________________________________________________
© 2014, IJIRAE- All Rights Reserved Page -256
Fig. 2 The Dataflow of DTBDM
First, we use isolation module to decide whether the pixel value is in a smooth region. If the result is negative, we conclude
that the current pixel belongs to noisy free. Otherwise, if the result is positive, it means that the current pixel might be a noisy
pixel or just situated on an edge. The fringe module is used to confirm the result. If the current pixel is situated on an edge, the
result of fringe module will be negative (noisy free); otherwise, the result will be positive. If isolation module and fringe module
cannot determine whether current pixel belongs to noisy free, the similarity module is used to decide the result. It compares the
similarity between current pixel and its neighboring pixels. If the result is positive, pi,j is a noisy pixel; otherwise, it is noise free.
The following sections describe the three modules in detail.
1) Isolation Module: The pixel values in a smooth region should be close or locally slightly varying, as shown . The differences
between its neighbouring pixel values are small. If there are noisy values, edges, or blocks in this region, the distribution of the
values is different, as shown in Figure. 3. Therefore, we determine whether current pixel is an isolation point by observing the
smoothness of its surrounding pixels. Figure. 4 shows an example of noisy image. The pixels with shadow suffering from noise
have low similarity with the neighbouring pixels and the so-called isolation point. The difference between it and its neighbouring
pixel value is large. According to the above concepts, we first detect the maximum and minimum luminance values in WTopHalf ,
named as TopHalf_max, TopHalf_min, and calculate the difference between them, named as TopHalf_diff. For WBottomHalf , we
apply the same idea to obtain BottomHalf_diff. The two difference values are compared with a threshold Th_IMa to decide
whether the surrounding region belongs to a smooth area. The equations are as
TopHalf_ diff = TopHalf_ max - TopHalf _min (3)
BottomHalf diff = BottomHalf max - BottomHalf_ min (4)
DecisionI = { true, if ( TopHalf diff ≥ Th _IMa)
or ( BottomHalf _dif f ≥ Th _IMa)
{ false, otherwise: (5)
Next, we take pi,j into consideration. Two values must be calculated first. One is the difference between f i,j and
TopHalf_max; the other is the difference between and f i,j and TopHalf_min. After the subtraction, a threshold ThIMb is used to
compare these two differences. The same method as in the case of WBottomHalf is applied. The equations are as
IM_TopHalf
{ True, if ( |f i,j - TopHalf_ max| ≥ Th _IMb)
= or | f i,j - TopHalf _min| ≥ Th_ IMb) (6)
{ false, otherwise.
IM _BottomHalf
= { true, if ( |f i,j _ BottomHalf_max| ≥ Th _IMb)

or ( | f i,j _ BottomHalf_ min| ≥ Th _IMb)
{ false, otherwise. (7)
________________________________________________________________________________________________________
{ true; if (IM _TopHalf = true)

DecisionII = or (IM_ Bottom_Half = true) (8)
{ false, otherwise.
Finally, we can make a temporary decision whether pi,j belongs to a suspected noisy pixel or is noisy free.
Fig. 3 A smooth and non smooth region in Lena
Fig. 4 The difference between noisy and neighbouring pixels in Lena
2) Fringe Module: If pi,j has a great difference with neighbouring pixels, it might be a noisy pixel or just situated on an edge,
as shown in Figure. 5. How to conclude that a pixel is noisy or situated on an edge is difficult. In order to deal with this case, we
define four directions, from E1 to E4, as shown in Figure. 6. We take direction E1 for example. By calculating the absolute
difference between f i;j and the other two pixel values along the same direction, respectively, we can determine whether there is an
edge or not. The detailed equations are as
FM _E1 = false; if (| a - fi,j| ≥Th _FMa)
or (|h – fi,j| ≥ Th_ FMa)
or (|a - h| ≥ Th _FMb)
= true, otherwise. (9)
FM _E2 = false; if (|c - fi,j | ≥ Th_ FMa)

or (|f -- fi,j | ≥ Th _FMa)
or(|c –f| ≥ Th _FMb)
= true; otherwise: (10)
FM_ E3 = false; if (|b - fi,j | ≥Th _FMa)
or |g - fi,j | ≥Th _FMa)
or (|b - g| ≥Th _FMb)
true; otherwise (11)
FM _E4 = false; if (|d - fi,j |≥ Th _FMa)
or (|e - fi,j | ≥ Th _FMa)

or (|d -e| ≥Th _FMb)
= true; otherwise (12)
DecisionIII = false; if (FM_ E1) or (FM_ E2)

or (FM E3) or (FM _E4)
= true; otherwise (13)
________________________________________________________________________________________________________
Fig. 5 The edge region in Lena Fig. 6 Four directions in DTBDM

3)Similarity Module: The last module is similarity module. The luminance values in mask W located in a noisy-free area might
be close. The median is always located in the center of the variational series, while the impulse is usually located near one of its
ends. Hence, if there are extreme big or small values, that implies the possibility of noisy signals. According to this concept, we sort
nine values in ascending order and obtain the fourth, fifth, and sixth values which are close to the median in mask W. The fourth,
fifth, and sixth values are represented as 4thinWi;j, MedianInWi,j, and 6thinWi,j. We define Maxi,j and Mini,j as
Maxi,j = 6thinWi,j + Th_ SMa,
Mini,j = 4thinWi,j - Th _SMa: (14)
Maxi,j and Mini,j are used to determine the status of pixel pi,j. However, in order to make the decision more precisely, we do some
modifications as
Nmax = { Maxi,j, if (Maxi.j ≤ MedianInWi,j
+Th _SMb)
{ MedianInWi;j
+ Th _SMb, otherwise (15)
Nmin = { Mini,j if (Mini,j ≥MedianInWi;j

- Th _SMb)
{ MedianInWi,j
- Th_ SMb, otherwise: (16)
Finally, if f i,j is not between Nmax and Nmin, we conclude that pi;j is a noise pixel. Edge-preserving image filter will be
used to build the reconstructed value. Otherwise, the original value f i,j will be the output. The equation is as
DecisionIV = { true, if ( f i,j ≥ Nmax) or ( f i,j ≤ Nmin

{ false, otherwise.
Obviously, the threshold affects the quality of denoised images of the proposed method. A more appropriate threshold
contributes to achieve a better detection result. However, it is not easy to derive an optimal threshold through analytic formulation.
The fixed values of thresholds make our algorithm simple and suitable for hardware implementation. According to our extensive
experimental results, the thresholds Th_ IMa, Th_ IMb, Th _FMa, Th _FMb, Th _SMa, and Th_ SMb are all predefined values and
set as 20, 25, 40, 80, 15, and 60, respectively.
B Edge-Preserving Image Filter

To locate the edge existing in the current W, a simple edge-preserving technique which can be realized easily with VLSI
circuit is adopted. The dataflow of edge-preserving image filter is shown in Figure. 7. Here, we consider eight directional
differences, from D1 to D8, to reconstruct the noisy pixel value. Only those composed of noise-free pixels are taken into account
to avoid possible misdetection. Directions passing through the suspected pixels are discarded to reduce misdetection. Therefore,
we use Maxi,j and Mini,j, defined in similarity module, to determine whether the values of d, e, f, g, and h are likely corrupted,
respectively. If the pixel is likely being corrupted by noise, we don’t consider the direction including the suspected pixel. In the
second block, if d, e, f, g, and h are all suspected to be noisy pixels, and no edge can be processed, so f i,j (the estimated value of p
i,j) is equal to the weighted average of luminance values of three previously denoised pixels and calculated as (a + b x 2 + c) /4. In
other conditions, the edge filter calculates the directional differences of the chosen directions and locates the smallest one (Dmin)
among them in the third block.The equations are as follows:
D1 = |d - h| + |a -e|, D2 = |a –g| + |b –h|, D3 = |b – g| x 2, D4 = |b -f| + |c -g|, D5 = |c -d| + |e -f|, D6 = |d –e| x 2

D7 = |a – h| x 2,D8 =|c – f| x 2 (17)
________________________________________________________________________________________________________
Fig. 7 Dataflow of edge-preserving image filter & its eight directional differences
Also f î,j = (a + d + e + h)/4, if Dmin = D1;

(a + b + g + h)/4, if Dmin = D2;
(b + g)/2, if Dmin = D3;
(b + c + f + g) /4, if Dmin = D4;
(c + d + e + f)/4, if Dmin = D5;
(d +e)/2, if Dmin = D6;
(a + h)/2, if Dmin = D7;
(c + f)/2, if Dmin = D8. (18)
In the last block the smallest directional difference implies that it has the strongest spatial relation with pi,j, and probably there
exists an edge in its direction. Hence, the mean of luminance values of the pixels which possess the smallest directional difference
is treated as f î;j. After f î;j is determined, a tuning skill is used to filter the bias edge. If f î,j obtain the correct edge, it will
situate at the median of b, d, e, and g because of the spatial relation and the characteristic of edge preserving. Otherwise, the values
of f¯ i;j will be replaced by the median of four neighbouring pixels (b, d, e, and g). We can express f¯ i;j as
f¯ i;j = Median(f î;j, b, d, e, g) (19)
III. VLSI IMPLEMENTATION OF DTBDM
DTBDM has low computational complexity and requires only two line buffers instead of full images, so its cost of VLSI
implementation is low. For better timing performance, we adopt the pipelined architecture to produce an output at every clock
cycle .The architecture adopts an adaptive technology and consists of two main blocks:decision-tree-based impulse detector, edge-
preserving image filter. The proposed method employs a histogram equalization to improve the quality of reconstructed image.
The decision-tree-based impulse detector is composed of three modules (isolation module, fringe module, and similarity module).
Each of them is described in the following sections.
1)Isolation module:
The architecture of IM consists of a comparator CMPL which is used to output the larger value from the two input values
while the comparator CMPS is used to output the smaller value from the two input values. The first two-level comparators are used
to find TopHalf_max and TopHalf_min. The SUB unit is used to output the difference which is subtracted the lower input
(TopHalf_min) from the upper one (TopHalf_max), and the |SUB| unit is used to output the absolute value of difference of two
inputs. The GC is the greater comparator that will output logic 1 if the upper input value is greater than the lower one. The OR
gate is employed to generate the binary result for IM_TopHalf. Finally, if the result of Decision II is positive, pi;j might be a noisy
pixel or situate on an edge. The next module (FM) will be used to confirm the result.
Fig. 8. Architecture of IM
________________________________________________________________________________________________________
2 )Fringe Module
The architecture of FM is composed of four small modules, from FM_1 to FM_4, and each of them is used to determine
its direction, as mentioned in Section 2.1.2. Fig 5.2.2.2 is a detailed implementation of FM_1. Since E1 is the direction from a to h
, the relation between a, h, and f i,j must be referenced. The three |SUB| units are used to determine the absolute differences
between them. The GC is described in the above section and the NOR gate is used to generate the result of FM _E1. If the result is
positive, we consider that f i,j is on the edge E1 and regard it as noise free.
Fig 9 Architecture of FM Fig. 10 Architecture of FM_1 module

3 )Similarity Module:
If IM and FM can’t determine whether f i,j belongs to a noisefree value or not, SM is used to confirm the result. This
architecture that is designed to accelerate the, sorting speed to obtain the fourth value in mask W. The detailed implementation of
module M0 is also shown. If a is greater than b, C01 is set to 1; otherwise, C01 is set to 0. The eight GC units are used to
determine the values from C01 to C08. After comparing, a combined unit is used to combine the results of each comparator to
obtain a number between 0 and 8. The number indicates the order of value in mask W. If a is the smallest value in mask W, the
output of the M0 module is 0; if a is the biggest value in mask W, the output is 8. The architectures of other modules (M1 to M8)
are almost the same as M0, with only little difference.
Fig 11 Architecture of SM
4)Edge-Preserving Image Filter
The Edge-Preserving Image Filter is composed of two modules, minED generator and average generator (AG). The
architecture consists of minED generator which is used to determine the edge that has the smallest difference. Eight directional
differences are calculated with twelve |SUB|, four ADD, and four shifter units. Then, the smallest one is determined by using the
Min Tree unit. Min Tree is made up of a series of comparators. After that, the mean of luminance values of the pixels which
process the smallest directional difference (Dmin) can be obtained from the average generator.
Fig 12 Architecture of minED Generator & Average generator
________________________________________________________________________________________________________
As mentioned already if pi;j-1; pi;j+1; pi+1;j-1; pi+1;j and pi+1, j+1 are all suspected to be noisy pixels, the final MUX will output
(a +b x 2 +c)/4. Otherwise, the MUX will output the mean of the pixel values which process Dmin. Some directional differences
are determined according to four pixel values, so its reconstructive values also need four pixel values. Two-level ADD and shifter
units are used to complete our calculation. As for the chip implementation, the gate counts or silicon area of one multiplier or
division is much larger than one shifter. In our design, all multipliers or divisions will be replaced by shifter units in order to lower
the hardware cost. After above computations, we sort b, d, e, and g in order. The reconstructed value f î;j obtained from edge-
preserving filter will be compared with the second and third values, named as SortFour2, SortFour3, and the final value f¯ i;j is
obtained from the equation as
f¯ i;j = { SortFour2; if (SortFour2 > f î;j)

{ SortFour3; if (SortFour3 < f î;j)
{ f¯ i;j, otherwise. (20)
5) Histogram Equalization
Image enhancement is a mean as the improvement of an image appearance by increasing dominance of some features or by
decreasing ambiguity between different regions of the image. Image enhancement processes consist of a collection of techniques
that seek to improve the visual appearance of an image or to convert the image to a form better suited for analysis by a human or
machine Numerous enhancement methods have been proposed but the enhancement efficiency, computational requirements, noise
amplification, user intervention, and application suitability are the common factors to be considered when choosing from these
different methods for specific image processing application. Histogram processing is the act of altering an image by modifying its
histogram. Common uses of histogram processing include normalization by which one makes the histogram of an image as flat as
possible. This is also known as contrast enhancement. Intensity transformation functions based on information extracted from
image such as enhancement, compression, segmentation and description.
Histogram Equalization is a technique that generates a gray map which changes the histogram of an image and redistributing
all pixels values to be as close as possible to a user –specified desired histogram. HE allows for areas of lower local contrast to
gain a higher contrast. Histogram equalization automatically determines a transformation function seeking to produce an output
image with a uniform Histogram. Histogram equalization is a method in image processing of contrast adjustment using the image
histogram. This method usually increases the global contrast of many images, especially when the usable data of the image is
represented by close contrast values. Through this adjustment, the intensities can be better distributed on the histogram. Histogram
equalization accomplishes this by effectively spreading out the most frequent intensity values. Histogram equalization
automatically determines a transformation function seeking to produce an output image with a uniform Histogram.
IV. RESULTS AND OBSERVATION
A. Simulation Results
The performance analysis of all architectures are analysed and it is simulated using ModelSim simulator .
Fig.13 Simulation waveform of Decision Tree
________________________________________________________________________________________________________
Fig.14 Simulation waveform of DTBDM
Fig.15 Simulation waveform of DTBDM
a)Noisy image b)Denoised image c)histogram equalized image
Fig.16 Resultsof denoising method and image restoration with image enhancement
V. CONCLUSIONS
A low-cost VLSI architecture for efficient removal of random-valued impulse noise is proposed in this paper. The approach
uses the decision-tree-based detector to detect the noisy pixel and employs an effective design to locate the edge. With adaptive
skill, the quality of the reconstructed images is notable improved. Our extensive experimental results demonstrate that the
performance of our proposed technique is better than the previous lower complexity methods and is comparable to the higher
complexity methods in terms of both quantitative evaluation and visual quality. It requires only low computational complexity and
two line memory buffers. Therefore, it is very suitable to be applied to many real-time applications. Furthermore, histogram
equalization is used to enhance the effects of removal of impulse noise.
________________________________________________________________________________________________________
ACKNOWLEDGMENT
I am grateful to Mr.Bibin Binu Simon, Assistant Professor, Department of Electronics and Communication Engineering,
Mangalam College of Engineering, for his innovative ideas in proceeding with this proposed system.
REFERENCES
[1] Chih-Yuan Lien,Chien-Chuan Huang,Pei-Yin Chen, and Yi-Fan Lin “An Efficient Denoising Architecture for Removal of
Impulse Noise in
Images”,IEEE Transactions on computers,vol.62,no.4,April 2013.
[2] S. Zhang and M.A. Karim, “A New Impulse Detector for Switching Median Filter,” IEEE Signal Processing Letters, vol. 9, no.
11, pp Nov. 2002.
[3] R.H. Chan, C.W. Ho, and M. Nikolova, “Salt-and-Pepper Noise Removal by Median-Type Noise Detectors and Detail-
Preserving Regularization,”
IEEE Trans. Image Processing, vol. 14, no. 10, pp. 1479-1485, Oct. 2005..
[4] P.-Y. Chen and C.-Y. Lien, “An Efficient Edge-Preserving Algorithm for Removal of Salt-and-Pepper Noise,” IEEE Signal
Processing Letters,
vol. 15, pp. 833-836, Dec. 2008.
[5] T. Sun and Y. Neuvo, “Detail-Preserving Median Based Filters in Image Processing,” Pattern Recognition Letters, vol. 15, pp.
341-347, Apr. 1994.
[6] W. Luo, “An Efficient Detail-Preserving Approach for Removing Impulse Noise in Images,” IEEE Signal Processing Letters,
vol. 13, no. 7, pp. 413-416, July 2006.
________________________________________________________________________________________________________

Useful Info

Uploaded by

Copyright:

Available Formats

Useful Info

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Useful Info

Uploaded by

Copyright:

Available Formats

International Journal of Innovative Research in Advanced Engineering (IJIRAE) ISSN: 2349-2163

Volume 1 Issue 7 (August 2014 ) http://ijirae.com

An FPGA Based Image Denoising Architecture with Histogram

Keywords— FPGA, Image denoising, impulse noise, impulse detector, Xilinx

II. DECISION TREE BASED DENOISING METHOD

A .Decision-Tree-Based Impulse Detector

Fig. 1 A 3X 3mask centered on pi,j

Fig. 2 The Dataflow of DTBDM

= { true, if ( |f i,j _ BottomHalf_max| ≥ Th _IMb)

{ true; if (IM _TopHalf = true)

Fig. 3 A smooth and non smooth region in Lena

Fig. 4 The difference between noisy and neighbouring pixels in Lena

FM _E2 = false; if (|c - fi,j | ≥ Th_ FMa)

FM _E4 = false; if (|d - fi,j |≥ Th _FMa)

or (|e - fi,j | ≥ Th _FMa)

DecisionIII = false; if (FM_ E1) or (FM_ E2)

Fig. 5 The edge region in Lena Fig. 6 Four directions in DTBDM

Nmin = { Mini,j if (Mini,j ≥MedianInWi;j

DecisionIV = { true, if ( f i,j ≥ Nmax) or ( f i,j ≤ Nmin

B Edge-Preserving Image Filter

D1 = |d - h| + |a -e|, D2 = |a –g| + |b –h|, D3 = |b – g| x 2, D4 = |b -f| + |c -g|, D5 = |c -d| + |e -f|, D6 = |d –e| x 2

Also f ^i,j = (a + d + e + h)/4, if Dmin = D1;

Fig 9 Architecture of FM Fig. 10 Architecture of FM_1 module

Fig 12 Architecture of minED Generator & Average generator

f¯ i;j = { SortFour2; if (SortFour2 > f ^i;j)

IV. RESULTS AND OBSERVATION

Fig.13 Simulation waveform of Decision Tree

Fig.14 Simulation waveform of DTBDM

Fig.15 Simulation waveform of DTBDM

a)Noisy image b)Denoised image c)histogram equalized image

You might also like