of 16
8/22/2019 908502
1/16
IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 10, NO. 3, MARCH 2001 367
Transform-Based Image Enhancement Algorithmswith Performance Measure
Sos S. Agaian, Member, IEEE, Karen Panetta, Senior Member, IEEE, and Artyom M. Grigoryan, Member, IEEE
AbstractThis paper presents a new class of the frequencydomain-based signal/image enhancement algorithms includingmagnitude reduction, log-magnitude reduction, iterative mag-nitude and a log-reduction zonal magnitude technique. Thesealgorithms are described and applied for detection and visual-ization of objects within an image. The new technique is basedon the so-called sequency ordered orthogonal transforms, whichinclude the well-known Fourier, Hartley, cosine, and Hadamardtransforms, as well as new enhancement parametric operators.A wide range of image characteristics can be obtained from asingle transform, by varying the parameters of the operators.We also introduce a quantifying method to measure signal/imageenhancement called EME. This helps choose the best parametersand transform for each enhancement. A number of experimentalresults are presented to illustrate the performance of the proposedalgorithms.
Index TermsAlpha-rooting, detection, frequency domain en-hancement, magnitude-reduction, sequency ordered transforms,visualization.
I. INTRODUCTION
IT IS well-known that image enhancement is a problem-ori-
ented procedure. The goal of the image enhancement is to
improve the visual appearance of the image, or to provide a
better transform representation for future automated image
processing (analysis, detection, segmentation, and recognition).
Many methods have been proposed for image enhancement[11], [13], [14]. A survey of digital image enhancement
techniques can be found in [1], [34], [8], [26]. Most of those
methods are based on gray-level histogram modifications
[11], [12], while other methods are based on local contrast
transformation and edge analysis [14], [17], or the global
entropy transformation [25]. In all of these methods, there are
no general standards for image quality which could be used as
a design criteria for image enhancement algorithms.
At present, there is no general unifying theory of image
enhancement. Methods of image enhancement techniques
can be generally classified into two categories: spatial do-
main methods, which operate directly on pixels, including
Manuscript received August 12, 1999; revised November14, 2000. This workwas supported in part by NASA under Grant NAG8-1311. The associate editorcoordinating the review of this manuscript and approving it for publication wasDr. Henri Maitre.
S. S. Agaian is with the Division of Engineering, The University of Texas,San Antonio, TX 78249-0669 (e-mail: [email protected]).
K. Panetta is with the Department of Electrical Engineering and ComputerScience, Tufts University, Medford, MA 02155 (e-mail: [email protected]).
A. M. Grigoryan is with CAMDI Laboratory, Department of Electrical En-gineering, Texas A&M University, College Station, TX 77843-3128 (e-mail:[email protected]).
Publisher Item Identifier S 1057-7149(01)01662-1.
Fig. 1. Diagram of the image enhancement with C ( p ; s ) .
region-based and rational morphology based, and frequency
domain methods. These methods operate on transforms of the
image, such as the Fourier, wavelet, and cosine transforms. The
basic advantages of transform image enhancement techniques
are 1) low complexity of computations and 2) the critical roleof the orthogonal transforms in digital signal/image processing,
where they are used in different stages of processing such as
filtering, coding, recognition, and restoration analysis. Image
transforms give the spectral information about an image, by
decomposition of the image into spectral coefficients that
can be modified (linearly or nonlinearly), for the purposes of
enhancement and visualization. The resulting advantage is that
it is easy to view and manipulate the frequency composition of
the image, without direct reliance on spatial information.
In [8], a comparative analysis of transform based image en-
hancement techniques is given. It includes techniques such as
alpha-rooting, modified unsharp masking, and filtering, which
are all motivated by the human visual response. The analysis ofthe existing transform based image enhancement techniques [1],
[8], [13], [34]shows that there are the common problems which
need to be solved, because
1) such methods introducecertain artifacts (in [8] they called
these artifacts objectionable blocking effects);
2) such methods cannot simultaneously enhance all parts of
the image very well;
3) it is difficult to select optimal processing parameters,
and there is no efficient measure that can be served as a
building criterion for image enhancement.
Finding a solution to this problem is very important espe-
cially when the image enhancement procedure is used as a
preprocessing step for other image processing techniques suchas detection, recognition, and visualization. It is also important
when constructing an adaptive transform based image enhance-
ment technique. The research work presented here offers novel
frequency domain based image enhancement methods for ob-
ject detection and visualization. The new technique is based
on the so called sequency ordered [30] orthogonal transforms
such as Fourier, Hartley, cosine, and Hadamard transforms
and new enhancement operators. A new class of the fast
trigonometric systems is used for performing the transform
coefficients manipulation operations. A quantitative measure
10577149/01$10.00 2001 IEEE
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-8/22/2019 908502
2/16
368 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 10, NO. 3, MARCH 2001
Fig. 2. (a) Linear combination of the clock and (b) moon images, which results in (c) an illegible image.
Fig. 3. Enhancement of the original image (a) via -rooting based on the (b) Fourier, (c) Hadamard, and (d) cosine transforms when = 0 : 9 2 .
of image enhancement is introduced. The technique devel-
oped here has been successfully employed on NASAs Earth
Observing System satellite data products for the purpose of
anomaly detection and visualization. These satellites collect
a Tera-byte of data per day, and fast and efficient methods
are crucial for analyzing these data.
The paper is organized as follows: In Section I, the sequency-
ordered transformis briefly described, which will be used forthe
image enhancement. Then in Section II, we define the so-called
sequency ordered orthogonal transform, quantitative measure of
signal/image enhancement and, then, describe in general a trans-
form-based image enhancement algorithm. Our first technique
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
8/22/2019 908502
3/16
AGAIAN et al.: TRANSFORM-BASED IMAGE ENHANCEMENT ALGORITHMS WITH PERFORMANCE MEASURE 369
Fig. 4. The -rooting by the 2-D cosine (from the top to bottom), Hadamard,Fourier, and Hartley transforms.
combines the magnitude and log-magnitude reductions for en-
hancement. Section III focuses on a number of experiments in
order to evaluate the enhancement algorithm. Zonal magnitude
reduction methods are given in Section IV, wherein the com-
parative analysis of transforms based image enhancement algo-
rithms (theoretical and experimental results) are provided. Fi-
nally, in Section V, a discussion and some concluding remarks
are given.
II. BACKGROUND
A. Sequency Ordered Systems
In this section, we review frequency ordered systems, de-
scribe their properties, and introduce a new class of nonsinu-
soidal sequency ordered systems. The latter can be implemented
with low computational complexity.
When analyzing signals and systems, it is useful to map data
from the time domain into another domain (in our case, the fre-
quency domain). The basic characteristics of a complex wave
are the amplitude and phase spectra. Specifying amplitude and
phase spectra is an important concept for complex waves. For
example, an amplitude spectrum contains information about the
energy content of a signal and the distribution of the energy
among the different frequencies, which is often used in many
applications. To achieve this, the real variable, , is generalizedtothe complex variable, , which thenismapped back via
inverse mapping. For example, the Fourier transform maps the
real line (time domain) into the complex plane, or real wave into
the complex one. This, however, requires a high complexity in
implementation, since it involves complex multiplications and
additions.
It is obvious to ask the question: Is it possible to construct a
discrete orthogonal system which maps a real signal to another
real signal while maintaining the advantages of a complex do-
main? The motivation for this question is that it is often easier
and efficient, especially from the standpoint of calculation, to
deal with real rather than complex numbers. If we accomplish
this task, we can obtain significant computational advantages in
signal processing, or more specifically, in the signal enhance-
ment.
The one-dimensional (1-D) discrete Fourier transform (1-D
DFT) is given by
(1)
The inverse 1-D discrete Fourier transform is defined as
(2)
The Hellers identity allows to rewrite the
Fourier transform pairs as
(3)
Thus, the 1-D Fourier transform maps the time domain signal
into the frequency domain. The sum of the cosine products can
be defined as the real components of the spectrum, and the
sum of the sine products can be defined as the imaginary com-
ponents of the spectrum. To compute these components, one can
use the known algorithms of the fast Fourier transform [1], [34]
or by using a new approach, an efficient manageable split al-
gorithm [27] for computing the Fourier and other unitary trans-
forms.
We now introduce a new system which has an inverse trans-
form as well as the basic advantages of the complex domain.
Definition 1: The rate at which a function crosses the
zero-axis is called its sequency (as an analog to frequency).
We now investigate the mapping systems, or transforms,
which meet the following properties of a special sequency-or-
dered system.
Definition 2: A special sequency-orderedfunction set is any
set of functions, which satisfies the following properties:
1) the transform can be represented in the form of
(4)
2) and are sequency-ordered functions. and
respectively can be considered as real and imagi-
nary components of the sums .
It is easy to see that the known orthogonal transforms such
as the Hartley, cosine, sine, and Hadamard transforms are the
particular cases of the sequency-ordered systems.
Remark 1: If and ,
, then the sequency-ordered system be-
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-8/22/2019 908502
4/16
370 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 10, NO. 3, MARCH 2001
Fig. 5. Enhanced images via the -rooting based on the (a)(c) Fourier transform and (d)(f) Hadamard transform.
comes the discrete Hartley transform of a 1-D, discrete real
function, , is defined as [3]
(5)
where . The Hartley transform is sim-
ilar to the Fourier transform, but only generates real coefficients
rather than complex ones.
Remark 2: If , ,
and , , then the se-
quency-ordered system becomes the cosine transform. Really,
the discrete cosine transform is determined by the basis func-tions
if
if
(6)
(7)
Remark 3: If and , if is
even, and , if is odd, then the sequency-or-
dered system becomes the CalSal WalshHadamard transform
(C-SWHT). Really, the C-SWHT is defined as
(8)
where
, if ;
, if ;
denotes the sequency.
The system is the set of the Walsh ordered functions,
(9)
(10)
Remark 4: A transform definition via a parametric class of
trigonometric systems is
(11)
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
http://-/?-http://-/?-8/22/2019 908502
5/16
AGAIAN et al.: TRANSFORM-BASED IMAGE ENHANCEMENT ALGORITHMS WITH PERFORMANCE MEASURE 371
(a)
(b)
Fig. 6. Enhancement by C ( p ; s ) coefficient and Fourier transform. (a) Original image and (b) enhanced images.
where
(12)
for some constants , and .
Properties of these kinds of systems, including the fast algo-
rithms, can be found in [6].
Remark 5: A new class of nonsinusoidal function transforms
is defined via a parametric class of trigonometric systems as
sgn sgn
(13)
for some constants , and .
Similar to the Fourier transform, one can define the magni-
tude and phase of the real transform . The phase asso-
ciated with is defined as
(14)
where and are respectively the sum of the real and
imaginary components in (11). The magnitude is defined as
(15)
The power
(16)
and phase spectra can be recombined to reconstruct completely
the .
Given an image of sizes , we consider a two-
dimensional (2-D) unitary transform
(17)
where is the set of basis functionsof the transform , and ,
is a complete set of orthogonal functions. and are
coefficients of the transform.
It is clear that the magnitude of the sequency-ordered sys-
tems are similar to the magnitude of the Fourier transform. This
fact points to possibility of construction unified transform based
enhancement algorithms for all sequency-ordered systems.
In the next section, we will show that the above defined mag-
nitude information provides useful information for object loca-
tion.
B. General Transform-Based Image Enhancement Algorithm
Analyzing the existing transform-based enhancement algo-
rithms ( -rooting andmagnitude reductionmethods[2], [8]), we
find a common algorithm, which encompasses all of these tech-
niques. The actual procedure of the signal/image enhancement
via an invertible transform consists of the following three steps:
Step 1) perform the orthogonal transform;
Step 2) multiply the transform coefficients, and ,
by some factor, ;
Step 3) perform the inverse orthogonal transform.
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-8/22/2019 908502
6/16
372 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 10, NO. 3, MARCH 2001
(a) (b)
(c) (d)
Fig. 7. Measure of log-enhancement by Fourier, Hadamard, and cosine transforms. (a) Fourier enhancement, (b) differenceE M E
(
;
)
0 E M E
(
;
)
, (c)difference E M E ( ; ) 0 E M E ( ; ) , and (d) difference E M E ( ; ) 0 E M E ( ; ) .
The frequency ordered system-based method can be repre-
sented as
(18)
where is an operator which could be applied on the com-
bination of and (particularly, on the modules of
the transform coefficients) or could be applied directly to
these coefficients. For instance, they could be , ,
or , . Basically, we are interested in thecases, when is an operator of magnitude (see cases
14, below) and when is performed separately on the
coefficients.
Let bethetransformcoefficients and let the enhance-
ment operator be of the form , where the
latter is a real function of the magnitude of the coefficients, i.e.,
. must be real because we only
wish to alter the magnitude information, not the phase informa-
tion. In the framework of this constraint, we have several possi-
bilities for , which can offer far greater flexibility:
1) c o n s t a n t (when the enhancement
preserves all constant information);
2) , (which is
the so-called modified -rooting [8]);
3) , , [2];
4) .
Denoting by the phase of the transform coeffi-
cient , we can write
(19)
where is the magnitude of the coefficients. Rather than
apply the enhancement operator directly on the transformcoefficients , we will investigate the operator which is
applied on the modules of the transform coefficients,
(20)
We assume the enhancement operator takes one
of the forms , , at every point
.
In practice, the coefficient is used in
for image enhancement. The optimal value of is image de-
pendent and should be adjusted interactively by the user [8]. For
simplicity of our reasonings, we will assume that in definition
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-8/22/2019 908502
7/16
AGAIAN et al.: TRANSFORM-BASED IMAGE ENHANCEMENT ALGORITHMS WITH PERFORMANCE MEASURE 373
Fig. 8. (a) Original image and (b)(d) 2-D Fourier transform enhancements when operating with C coefficients for ( ; ) equal respectively to (0.05, 0.05),(1.9, 0.05), (0.05, 0.9), and (1.9, 0.9).
of the coefficients we have . One
can ask: What are the optimal values of , , and ? Can one
choose , , and automatically? What is the best enhance-
ment frequency ordered system? What is the optimal size of the
transform, ?
Remark 6: The above approach can be used 1) on the whole
image, or via blockwise processing with block sizes 8, 16, 32,
and 64 and 2) on some low-pass or high-pass filtered image.
As an example, one can see in Fig. 1 that an original imagecan be divided first into a low-pass image and high-pass
image . The high-pass image is enhanced by multiplication
by and then recombined with the low-pass image [see
also [26], [8], when using the coefficient ].
C. Performance Measure of Enhancement
In this section, we present a new quantitative measure of
image enhancement.
The improvement in images after enhancement is often very
difficult to measure. A processed image can be said to have been
enhanced over the original image if it allows the observer to
better perceive the desirable information in the imaging. In im-
ages, the improved perception is difficult to qualify. There is
no universal measure which can specify both the objective and
subjective validity of the enhancement method [16]. In practice,
many definitions of the contrast measure are used [12], [16],
[17]. For example, the local contrast proposed by Gordon and
Rangayan was defined by the mean gray values in two rectan-
gular windows centered on a current pixel. Baghdan andNegrate
[17] proposed another definition of the local contrast based onthe local edge information of the image, in order to improve
the first mentioned definition. In [17], the local contrast method
proposed by Beghdadi and Negrate has been adopted, in order
to define a performance measure of enhancement. Use of sta-
tistical measures of gray level distribution measures of local
contrast enhancement (for example, mean, variance or entropy)
have not been particularly meaningful for mammogram images.
A number of images, which clearly illustrated an improved con-
trast, showed no consistency, as a class, when using these sta-
tistical measurements. A measure proposed in [12], which has
greater consistency than the statistical measures, is based on the
contrast histogram.
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-8/22/2019 908502
8/16
374 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 10, NO. 3, MARCH 2001
(a)
(b)
(c) (d)
Fig. 9. Fourier enhancement via log-reduction when coefficients C ( p ; s ) are calculated for one fixed parameter. (a) Surface of the enhancement measure ( =0 : 8 ), (b) image of the enhancement measure, (c) surface of the enhancement ( = 0 : 8 ), and (d) surface of the enhancement measure ( = 0 : 9 ).
Intuitively, it seems reasonable to expert that a image
enhancement measure values at given pixels should depend
strongly on the values at pixels that are close by weekly
on those that are further away and also this measure has to
related with human visual system. In our definition, we use a
modification of Webers and Fechners laws. In [31], Weber
established a visual law, argued that the human visual detection
depends on the ratio, rather than difference, between the light
intensity value and . The Weber
definition of contrast was used to measure the local contrast of
a single object. (One usually assumes a large background with
a small test object, in which case the average luminance will be
close to the background luminance. If there are many objects
this assumption do not hold.) Fechners law [32] proposed the
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
http://-/?-http://-/?-http://-/?-http://-/?-8/22/2019 908502
9/16
AGAIAN et al.: TRANSFORM-BASED IMAGE ENHANCEMENT ALGORITHMS WITH PERFORMANCE MEASURE 375
following relationship between the light intensity and
brightness:
(21)
where is a constant, and are the absolute
threshold and upper threshold of the human eye [33].Below, a new quantitative measure of image enhancement is
presented.
Let an image be split into blocks of
sizes , and let , , and are fixed enhancement parame-
ters (or, vector parameter). For a given class of orthogonal
transforms, we define a value as
follows:
(22)
(23)
where and respectively are the min-
imum and maximum of the image inside the block
, after processing the block by transform based enhance-
ment algorithm. The function is the sign function, ,
or , depending on the method of enhancement under
the consideration. The decision to add this function has been
done after studying various examples of enhancement by trans-
form methods using the different coefficients ,
. This will be demonstrated in the following sections.
Definition 3: is called a measure of enhancement, or
measure of improvement.
Definition 4: The best (optimal) transform relative to the
measure of enhancement is called a transform such that
. The image enhancement algorithm based
on this transform is called an optimal image improvement trans-
form-basedenhancement algorithm.
Selection of Parameters: Suppose the transform based en-
hancement algorithm depends on the parameters , and ,
or vector , i.e., .
Definition 5: Let be the best (optimal) transform. The
best (optimal) -transform-based enhancement image vector
parameter is called a parameter such that
.
It should be noted that the window size can be also includedin the vector as a parameter of optimal enhancement.
In the next section, the following problems are investigated.
How to design the best transform-based image enhancement
algorithm, and how to design the best -transform-based en-
hancement image vector parameter ?
III. EXPERIMENTAL RESULTS
In this section, we perform a number of experiments in order
to evaluate the enhancement algorithm for 1-D and 2-D signals.
For more clarity/visibility, we demonstrate the experimental re-
sults for 2-D signals such as the moon plus clock image.
(a)
(b)
Fig. 10. Hadamard enhancement via log-reduction when coefficientsC ( p ; s ) are calculated for fixed = 0 : 8 . (a) Surface of the enhancementmeasure ( = 0 : 8 ) and (b) image of the enhancement measure.
In our test cases, we use three classes of algorithms, namely,
the transform based enhancement algorithms via the operators
, , and respectively. We also present
the above mentioned iterative algorithm. For each of the cases,we present two classes of experiments. The first class shows
how to choose the best operator parameter (or, the best enhance-
ment algorithm) forthe given transform. Thesecond class shows
how to choose the best image enhancement transform for the
given image. A quantitative comparison of the methods is also
presented below.
In order to enhance our images before passing them through
a visualization algorithm, we reduce the magnitude informa-
tion of the image while leaving the phase information intact.
Since the phase information is much more significant than the
magnitude information in the determination of edges, reducing
the magnitude produces better edge detection capabilities. This
method also tends to reduce the low-frequency componentsrather than the high-frequency components (both the low-fre-
quency components, which are associated with sharp edges,
and high-frequency components, which are associated with the
edge elements).
The clock image was taken as the original, , and
then the moon image was superimposed with it. This results
in an illegible image, as shown in Fig. 2. The result, , is
an enhanced image, which can now be passed through a visual-
ization algorithm.
Case 1: Transform based enhancement algorithm via oper-
ator .
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
http://-/?-http://-/?-8/22/2019 908502
10/16
376 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 10, NO. 3, MARCH 2001
Fig. 11. Two-dimensional Fourier enhancement of the image (a) via the operator O with coefficients C , (b), and C , (c), when = 0 : 8 , = 1 : 5 , and = 0 : 8 ,and the histograms (d)(f) of the images (a)(c), respectively.
Test 1.1Choosing the Best Operator Parameter: This case
is known as modified -rooting or root filtering [8]. When
equals to zero, only the phase is retained. When , the
amplitude of the large transform coefficients are reduced
relative to the amplitude of the small coefficients, and the
result is enhanced edges and details in the image. Since most of
the edge information is contained in the high-frequency region
of the spectrum, the edges are enhanced by this method. By
varying the -level of the reduced image, we are able to enhance
the quality of our images for the visualization. Fig. 3(a)(c),
illustrates the process of enhancement of the image when the
parameter and the Fourier, Hadamard, and cosinetransforms are used. As we see in the above examples, the
magnitude reduction using servedto sharpen the image
as well as even out the brightness throughout the image. The
results of the visualization algorithms will be more accurate
because they will be operating on these enhanced images.
They will also be less dependent on magnitude variations
based on magnification and blurring, therefore making it much
easier to set a thresholding constant, which need not change
from image to image.
Test 1.2Choosing the Best Image Enhancement Transform
for the Given Image: Let be identical to after
the normalization by a constant.
The enhancement measure of the original image shown in
Fig. 2 is 4.5, or , where is the identical
transform. Fig. 4 shows four curves which describe the mea-
sure of the enhancement, when applying the Fourier, Hadamard,
cosine, and Hartley transforms. We see that on the whole in-
terval, where varies, the maximal measure of enhancement is
provided mostly by the cosine and Hadamard transforms. The
curves have two maximums, at points and ,
where the maximum measure is provided by the Fourier trans-
form (The best transform among the above transforms). The ex-
perimental results show that theparameter corresponds to the
best visual estimation of enhancement. The enhancement by thetransforms are very close between these two extreme points.
Case 2: Transform based enhancement algorithm via oper-
ator .
Test 2.1Choosing the Best Operator Parameter: Fig. 5 il-
lustrates the enhanced images by varying parameter , when
using the Fourier transform, (a)(c), and the Hadamard, (d)(f).
The log-magnitude reduction using serves to enhance
the edges around regions in the image. Fig. 6 demonstrates the
practical application of the proposed method on the images ob-
tained by NASAs Earth Observing System satellites.
Test 2.2Choosing the Best Image Enhancement Transform
for the Given Image: Fig. 7 illustrates the measure of the
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
http://-/?-http://-/?-8/22/2019 908502
11/16
AGAIAN et al.: TRANSFORM-BASED IMAGE ENHANCEMENT ALGORITHMS WITH PERFORMANCE MEASURE 377
Fig. 12. Two-dimensional Fourier transform enhancement of (a) the original image and (b)(d) results of the enhancement, respectively, for the coefficients C ,C , and C with = 0 : 9 , = 1 : 5 , and = 0 : 8 .
image enhancement by using different transforms and varying
parameters and respectively in the intervals and
. Fig. 7(a) shows the surface of the measure for the Fourier
method and (c) and (d) show the differences between the
measures when the Fourier, Hadamard, and cosine transforms
are used for enhancement. The results of the Fourier transform
based image enhancement are shown in Fig. 8, for the boundary
parameters. The large values of lead to the elimination of the
higher frequencies on the image spectrum, and the operatorworks as the filter of low frequencies. Contrarily, the small
values of increase the image enhancement.
Test 2.3Comparison: To illustrate the above method of en-
hancement, consider Fig. 9. Fig. 9 depicts the surfaces of mea-
sure , when one of the parameters is fixed.
Case 3: Transform based enhancement algorithm via oper-
ator .
Test 3.1Choosing the Best Operator Param-
eter: Combining the magnitude reduction and magnitude
reduction methods in accomplishes both the
sharpening and edge enhancements for a given image. In our
experiments, we found with and toFig. 13. Curves of the Fourier enhancement (a), (b), and (c), by two zones,when using the coefficients C , C , and C , respectively.
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
8/22/2019 908502
12/16
378 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 10, NO. 3, MARCH 2001
Fig. 14. Curve of the Fourier enhancement by (a) two zones and (b)(d) results of enhancement when radius of the first zone is 32, 64, and 127, respectively.
be the optimal magnitude reduction operator on the image.
Fig. 9 illustrates the surface of the enhancement measure for
and , when the Fourier transform based
enhancement image algorithm is used. Fig. 10 illustrates the
surface of the enhancement measure for , when the
similar Hadamard transform algorithm is applied.
Test 3.2Choosing the Best Image Enhancement Trans-
form: We face the problem of selecting the optimal orthogonal
transform for our application. Since our goal is to achieve
maximum accuracy in the detection of regions of interest aswell maximize computational speed, we must balance these
two factors and make a selection that is appropriate for our
application. Therefore, we analyzed the quality of the results
and the execution time for each of these orthogonal transform
algorithms.
Test 3.3Comparison: An enlarged example of the pro-
posed optimal magnitude reduction is shown in Fig. 11(a),
when using the Fourier transform (b) and comparing with
-rooting (c). The histograms (d)(f) show how the range
of intensities differ, when using the different coefficients for
the enhancement. The measure of enhancement is 9.84 and
7.80 when using respectively the coefficients and for
enhancement of the image. Fig. 12 illustrates for comparison
the outputs of the 2-D Fourier transform enhancement for all
three methods under consideration. One can see that the max-
imum measure of enhancement and best visual estimate occurs
when using the coefficient . All further references
to the magnitude reduction algorithm will be to this specific
combination of magnitude reductions.
It should be noted, that from standpoint of information theory,
the probability distribution, which conveys the most informa-
tion, is perfectly uniform [11]. Therefore, if we could obtain asuniform a histogram as possible, the image information could
be maximized.
A. Iterative Transform Based Image Enhancement Algorithm
We now discuss briefly an algorithm via the magnitude re-
duction approach. Naturally, one wonders if it is possible to fur-
ther enhance the image. Since the proposed magnitude reduc-
tion method works well at enhancing the edge information of
the image, a second pass through the magnitude reduction al-
gorithm might serve to further enhance the image. We have im-
plemented this iterative enhancement method and experimented
with it using various magnitude reduction coefficients, in order
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
http://-/?-http://-/?-8/22/2019 908502
13/16
AGAIAN et al.: TRANSFORM-BASED IMAGE ENHANCEMENT ALGORITHMS WITH PERFORMANCE MEASURE 379
Fig. 15. Two-dimensional Fourier enhancement for two zones ( = 1 6 ).
to find the optimal values for this configuration. The goal is to
find the optimal iteration parameters. We propose the followingalgorithm for iterative transform based image enhancement:
1) perform the orthogonal transform;
2) multiply the transform coefficients by some factors
3) repeat step 2 for several iterations, while varying coeffi-
cients
4) perform the inverse orthogonal transform.
IV. ZONAL TRANSFORM BASED ENHANCEMENT METHODS
The classical transform based image enhancement tech-
niques are performed uniformly over the entire frequencyspectrum. As an expansion of these kinds of techniques, we
propose varying the transform based image enhancement within
radially concentric zones. The motivation of using zones comes
from the fact that: a very short transform coefficient length
corresponds to the homogenous image blocks, a medium
transform coefficient length corresponds to the texture, and a
long transform coefficient length corresponds to the highly
active image blocks. By using this method, we achieve much
more flexibility and control over the magnitude reductions
in different regions within the frequency domain. By using
increasingly greater reductions in higher frequencies, we
manage to attenuate the high-frequency noise component of the
image. At the same time, we also maintain the edge enhancing
effects of the magnitude reduction algorithm.We now demonstrate zonal transform based image enhance-
ment via a few examples. For this, we describe the transform
based image enhancement algorithm (including iterative algo-
rithm) with two and three zones. It should be noted that, in the
case considered above, we have used one zone.
In order to accomplish this method, we first have to find
the maximum and minimum values within the frequency do-
main data. Then, using these maximum and minimum points as
end-markers, we divide the frequency domain into regions based
on each points magnitude distance from themaximum and min-
imum. We set distance dividers between the maximum and min-
imum points, which divide the frequency domain into regions.
Each region has the specified magnitude reduction value, , andlog-magnitude reduction value, . We next determine the four
pairs of and values, as well as three values of the distance
to specify our magnitude reductions. These values can be deter-
mined by means of the enhancement measure.
Case 4: Transform based image enhancement algorithms
with two zones, and .
a) The enhancement operator can be defined by
if
if(24)
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
8/22/2019 908502
14/16
380 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 10, NO. 3, MARCH 2001
Fig. 16. Two-dimensional Hadamard enhancement for two zones ( = 1 6 ).
Remark 7: The zone has to be very small and coefficient
has to be near to zero or zero. b) The enhancementoperator can be defined by
if
if ,(25)
where is a constant (which can be the mean of all ,
), , where is the size of the
input signal, , , and are constants (and ).
As examples, Fig. 13 illustrates the curves of the enhance-
ment measure for images, by using two zones and the Fouriertransform based enhancement method for the different operators
, a), , b), and , c), for the parameters
, , and . The varying parameter for
the curves is the radius of the first zone by which the area of
the Fourier spectrum is divided; the second zone is the rest of
the area. Fig. 13 shows that when radius of the zone increases,
the measure of image enhancement grows faster when using the
coefficient , than .
Fig. 14 illustrates the example of the image enhancement by
two zones with varying radius-parameter . As can be observed
from the experimental results, the proposed algorithm effec-
tively enhances the overall contrast and sharpness of the test im-
ages. A lot of details, that could not been seen in the test image,
have been clearly revealed.c) The enhancement operator can be defined by
if
if ,(26)
where
magnitude of the transform image;,
thresholding operator;
very small constant (or, zero).
is defined as , where
and are real functions and is a coefficient (or, a zero-
frequency component). For instance, ,
when and .
Figs. 15 and 16 illustrate the examples of the image enhance-
ment by using the Fourier and Hadamard transforms and two
zones. The threshold is and parameter takes values
0.5, 0.6, 0.7, 0.8, and 1.
Case 5: Transform Based Image Enhancement Algorithms
with Three Zones.
The enhancement operator can be defined by
if
if
if
(27)
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
8/22/2019 908502
15/16
AGAIAN et al.: TRANSFORM-BASED IMAGE ENHANCEMENT ALGORITHMS WITH PERFORMANCE MEASURE 381
This is a typical contrast stretching transform, which has to be
applied in the frequency domain (see [13, pp. 235237]).
V. CONCLUSION
A new class of frequency domain based signal/image en-
hancement algorithms (magnitude reduction, log-magnitude re-duction, iterative magnitude, and log-reduction zonal magnitude
techniques) have been described and applied for detection and
visualization on objects within an image. The new techniques
are based on the so-called sequency ordered orthogonal trans-
forms, which include the well-known fast orthogonal Fourier,
Hartley, cosine, and Hadamard transforms, as well as new en-
hancement parametric operators.
We have improved upon the current magnitude reduction
techniques and developed an entirely novel method. The
wide range of characteristics can be obtained from a single
transform by varying enhancement parameters. A quantitative
measure of signal/image enhancement was presented, which
demonstrated the optimal method to automatically choose thebest parameters and transform. The proposed algorithms are
simple to apply and design, which makes them practical. A
number of experimental results were given which illustrate the
performance of these algorithms. The comparative analysis
of transforms based image enhancement algorithms has been
described, too. Lastly, the comparison of the Fourier transform
and Walsh, cosine and Hartley transforms was given. We find
that for a negligible tradeoff of accuracy, one can use the Walsh
transform to achieve significantly higher performance enhance-
ment. For our purposes, where speed is a major concern, the
proposed method turns out to be a dramatic improvement over
existing methods. We have also proposed the zonal transform
based image enhancement algorithms.
REFERENCES
[1] S. S. Agaian, Advances and Problems of Fast Orthogonal Transform forSignal/Image Processing Applications, pp. 146215, 1990.
[2] R. Kogan, S. S. Agaian, and K. P. Lentz, Visualization using rationalmorphology and zonalmagnitude-reduction, Proc. SPIE, vol. 3304, pp.153163, 1998.
[3] R. P. Millane, Analytic properties of the Hartley transform and theirapplications, Proc. IEEE, vol. 82, pp. 413428, Mar. 1994.
[4] D. F. Elliot and K. R. Rao, Fast Transforms: Algorithms, Analyses, Ap-plications. New York: Academic, 1983.
[5] K. G. Beauchamp, Walsh Functions and Their Applications. NewYork: Academic, 1975.
[6] K. O. Egiazarian, S. S. Agaian, and J. Astola, On a parametric familyof discrete trigonometric transforms, Proc. SPIE, 1996.
[7] J. H. McClellan, Artifacts in alpha-rooting of images, in Proc. IEEEInt. Conf. Acoustics, Speech, Signal Processing, Apr. 1980, pp. 449452.
[8] S. Aghagolzadeh and O. K. Ersoy, Transform image enhancement,Opt. Eng., vol. 31, pp. 614626, Mar. 1992.
[9] R. Rajcsy and Liberman, Texture grandiuts as a depth one, Comput.Graph. Image Process., vol. 5, pp. 5257, 1976.
[10] D. L. Donoho, De-noising by soft-thresholding, IEEE Trans. Inform.Theory, vol. 41, pp. 613627, May 1995.
[11] A. Rosenfeld and A. C. Kak, Digital Picture Processing. New York:Academic, 1982, vol. 1.
[12] W. M. Morrow, R. B. Paranjape, R. M. Rangayyan, and J. E. L. De-sautels, Region-based contrast enhancement of mammograms, IEEETrans. Med. Imag., vol. 11, pp. 392406, Sept. 1992.
[13] A.Jain, Fundamentals of Digital Image Processing. Englewood Cliffs,NJ: Prentice-Hall, 1989.
[14] T. L. Ji, M. K. Sundareshan, and H. Roehrig, Adaptive image con-trast enhancement basedon human visual properties,IEEE Trans. Med.
Imag., vol. 13, pp. 573586, Dec. 1994.[15] N. Netravali and Presada, Adaptive quantization of picture signals
using spatial masking, Proc. IEEE, vol. 65, no. 4, pp. 536548, 1977.[16] J. K. Kim, J. M. Park, K. S. Song, and H.W. Park, Adaptive mammo-
graphic image enhancement using first derivative and local statistics,IEEE Trans. Med. Imag., vol. 16, pp. 495502, Oct. 1997.
[17] A. Beghcladiand A. L. Negrate, Contrast enhancement techniquebased
on local detection of edges, Comput. Vis., Graph., Image Process., vol.46, pp. 162274, 1989.[18] D.-C. Chang and W.-R. Wu, Image contrast enhancement based on a
histogramtransformationof localstandarddeviation,IEEE Trans. Med.Imag., vol. 17, pp. 518531, Aug. 1998.
[19] G. X. Ritter and J. N. Wilson, Handbook of Computer Vision Algorithmsin Image Algebra. Boca Raton, FL: CRC Press, 1996.
[20] J. B. Zimmerman, S. M. Pizer, E. V. Staab, J. R. Perry, W. McCartney,and B. C. Brenton, An evaluation of the effectiveness of adaptivehistogram equalization for contrast enhancement, IEEE Trans. Med.
Imag., vol. 7, pp. 304312, Dec. 1988.[21] R. Gonzalez and Wintz, Digital Image Processing, 2nd ed. Reading,
MA: Addison-Wesley, 1987.[22] D. Ballard and C. Brown, Computer Vision. Englewood Cliffs, NJ:
Prentice-Hall, 1982.[23] J. Harris, Constant variance enhancement: A digital processing tech-
nique, Appl. Opt., vol. 16, pp. 12681271, May 1977.
[24] P. Natendra and Rithch, Real-time adaptive contrast enhancement,IEEE Trans. Pattern Anal. Machine Intell., vol. PAMI-3, pp. 655661,Nov. 1981.
[25] A. Khellaf, A. Beghdadi, and H. Dupoiset, Entropic contrast enhance-ment, IEEE Trans. Med. Imag., vol. 10, pp. 589592, Dec. 1991.
[26] D. Wang and A. H. Vagnucci, Digital image enhancement, Comput.Vis., Graph., Image Process., vol. 24, pp. 363381, 1981.
[27] A.M. Grigoryanand S. S. Agaian, Split manageable efficient algorithmfor Fourier and Hadamard transforms, IEEE Trans. Signal Processing,vol. 48, pp. 172183, Jan. 2000.
[28] J. A. Saghri, P. S. Cheatham, and H. Habibi, Image quality measurebased on a human visual system model, Opt. Eng., vol. 28, no. 7, 1989.
[29] H. S. How and D. R. Tretter, Frequency characterization of the discretecosine transform, Proc. SPIE, vol. 1349, pp. 3142, 1990.
[30] H. F. Harmuth, Applications of Walsh function in communications,IEEE Spectrum, pp. 8242, 1969.
[31] I. E. Gordon, Theory of Visual Perception. New York: Wiley, 1989.
[32] G. T. Fechner, Elements of Psychophysics. New York: Rinehart &Winston, 1960, vol. 1.[33] L. E. Krueger, Reconciling Fechner and Stevens: Toward an unified
psychophysical law, Behav. Brain Sci., vol. 12, pp. 251320, 1989.[34] S. S. Agaian, Advances and Problems of Fast Orthogonal Transform
for Signal/Image Processing Applications. Moscow, Russia: Nauka,1991, pp. 99145.
Sos S. Agaian (M85) received the M.S. degree inmathematics and mechanics from Yerevan State Uni-versity, Armenia, the Ph.D. degree in mathematics
and physics from Steklov Institute of Mathematics,Academy of Sciences, Moscow, Russia, and theDoctor of Industrial Sciences degree from ComputerCenter, Academy of Sciences, USSR in 1985. Healso received the Full Professor Diploma from theSupreme Attestation Board, USSR, in 1986.
He wasa VisitingProfessorwith theDepartment ofElectrical Engineering and Computer Science, Tufts
University, Medford, MA, from 1993 to 1997. He was Senior Scientist withAWARE, Inc., from 1996 to 1997. He erved as a Chairman of the Departmentof Digital Signal Processing, National Academy of Sciences of Armenia, from1979 to 1993. In 1997, he joined the Division of Engineering, University ofTexas, San Antonio, where he is currently an Associate Professor. He is the au-thor of three books and 11 inventions and has written more than 185 papers onorthogonal and logical transforms, with application on compression, filtering,and recognition. His current research interests include signal and image pro-cessing, computer vision, visual communication, and applied mathematics.
Authorized licensed use limited to: Telecom ParisTech. Downloaded on July 26,2010 at 07:57:21 UTC from IEEE Xplore. Restrictions apply.
http://-/?-http://-/?-8/22/2019 908502
16/16
382 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 10, NO. 3, MARCH 2001
Karen Panetta (SM95) received the B.S. degreein computer engineering from Boston University,Boston, MA, and the M.S. and Ph.D. degrees inelectrical engineering from Northeastern University,Boston.
She is an Assistant Professor of electrical engi-neering and computer science at Tufts University,Medford, MA. Her research interests include visual-ization of complex data sets, fault simulation, large
system simulation, and behavioral modeling. Sheis currently a NASA JOVE Fellow for the NASALangley Research Center. Her research has been supported by Compaq, Intel,Analog Devices and, most recently, the NSF CAREER award. She holds twopatents in developing discrete-event simulation methodologies and algorithms.
Dr. Panetta is Faculty Advisor for the student chapter of IEEE at Tufts Uni-versity and a Member of ACM and the Society for Computer Simulation.
Artyom M. Grigoryan (S78M99) received the M.S degrees in mathematicsfrom Yerevan State University, Armenia, in 1978, imaging science fromMoscow Institute of Physics and Technology, Moscow, Russia, in 1980,and electrical engineering from Texas A&M University, College Station, in1999, and the Ph.D. degree in mathematics and physics from Yerevan StateUniversity, Armenia, in 1990.
From 1990to 1996,he wasa SeniorResearcherwith theDepartment ofSignaland Image Processing, Institute for Problems of Informatics and Automation,and Yerevan State University, Academy Science of Armenia. In 1996, he joined
the Department of Electrical Engineering, Texas A&M University, where he iscurrently a Research Engineer. He holds one patent in the development of theautomated 3-D fluorescent in situ hybridization spot counting on tissue microar-rays. He is author of 40 papers and specializes in the design of robust nonlinearand linear optimal filters, linear filtration, theory of fast one-dimensional andmultidimensional unitary transforms, and processing the fluorescent images.