GENERALIZED EQUALIZATION MODEL

FOR GLOBAL IMAGE CONTRAST

MAPPING
K.Mounika M.Sagar
M.Tech, PG student Associate Professor
[email protected] sagarmettu34 @gmail.com

ABSTRACT: a generalized leveling model for displays. Thus tone mapping techniques, e.g. the
image improvement. supported our analysis on the wide identified gamma correction, are used to
relationships between image bar chart and transfer the image into an acceptable dynamic vary.
distinction improvement white reconciliation, we a lot of refined tone mapping algorithms were
have a tendency to initial establish a generalized developed through the years, simply to call a few.
leveling model desegregation distinction Generally, tone mapping algorithms can be
improvement and white reconciliation into a classified into two categories by their
unified framework of protrusive programming of functionalities throughout the imaging method. A.
image bar chart. We have a tendency to show that White Balancing
several image improvement tasks may be
Because of the undesirable luminance or the
accomplished by the projected model exploitation
physical limitations of cheap imaging sensors, the
totally different configurations of parameters. With
captured image might carry obvious color bias.1To
2 shaping properties of bar chart remodel,
calibrate the color bias of image, we want to
specifically distinction gain and nonlinearity, the
estimate the worth of light source, the matter of
model parameters for various improvement
that known as constancy. Mistreatment suitable
applications may be optimized. We have a tendency
physical imaging model, one will get associate
to then derive associate degree best image
approximated brightness, so a linear trans form are
improvement formula that in theory achieves the
often applied to map the first image into an ideal
most effective joint distinction improvement and
one.
white reconciliation result with trading-off between
distinction improvement and tonal distortion.

I. INTRODUCTION

With the quick advance of technologies and

therefore the prevalence of imaging devices,
billions of digital pictures square measure being
created on a daily basis. Because of undesirable
source of illumination, unfavorable weather or
failure of the imaging device itself, the contrast and
tone of the captured image might not invariably be
satisfactory. Therefore, image improvement is
usually needed for each the aesthetic and pragmatic Fig.1. (a) is the illustration of traditional image
functions. In fact, image enhancement algorithms enhancement strategy, (b) is the illustration of
have already been wide applied in imaging devices joint image enhancement strategy
for tone mapping. As an example, in a typical
photographic camera, the CCD or CMOS array B. Distinction Enhancement
receives the photons passing through lens and then
Distinction improvement algorithms square
the charge levels square measure remodeled to the
measure wide used for the restoration of degraded
first image. Usually, the first image is kept in RAW
media, among that international bar chart leveling
format, with a bit-length too huge for traditional
is that the most popular choice as shown in Fig.1. integrating contrast enhancement and white
Alternative variants include native bar chart balancing into a unified framework of convex
equalization and therefore the abstraction filtering programming of image histogram. The enhanced
sort of strategies. As an example, in the half filter system flow is shown below. The processes carried
issued to market the variances of texture thus on out here are
enhance the image. In, a texture synthesis primarily
based formula is proposed for degraded media, like  RGP Panel Splitting
recent photos or films. On the other hand, remodel
 White Balancing
primarily based strategies additionally exist, e.g.
Curve let based algorithm associate adaptive  FFT & IFT
steering regression kernels planned to combine
image sharpening with de-noising. Despite of the  Bi-log Transform
plentiful literature on image improvement,
including those representatives listed on top of, 2  RGP restoration and
challenging problems for image improvement
 NTSC
square measure still not resolved first, how to reach
distinction improvement whereas conserving an White Balancing: White balance (WB) is the
honest tone. process of removing unrealistic color casts, so that
objects which appear white in person are rendered
II. PROPOSED SCHEME white in your photo. Proper camera white balance
has to take into account the "color temperature" of
A. Image Quality and Preserving Color space for
a light source, which refers to the relative warmth
Human Visual System The blog transform
or coolness of white light. Our eyes are very good
smoothly modifies the gradient of the
at judging what is white under different light
transformation so that in the region near zero it
sources, but digital cameras often have great
remains finite. A single constant C is provided to
difficulty with auto white balance (AWB), and can
tune this behavior, so as to adjust the meaning of
create unsightly blue, orange, or even green color
―region near zero‖. The default value of this
casts. Understanding digital white balance can help
constant is 1/ ln(10); this gives a unity transfer
you avoid these color casts, thereby improving
function at zero but other values can be applied as
your photos under a wider range of lighting
wished, to focus into the region near zero or not.
conditions.
The modified logarithmic transformation called
Bilog transform can be both one-sided and FFT: The Fourier transform is a representation of
symmetric, and thus can transform negative data to an image as a sum of complex exponentials of
scaled negative data. It can be applied to both the X varying magnitudes, frequencies, and phases. The
and Y data, when it becomes a bi-symmetric log Fourier transform plays a critical role in a broad
transform. Applying log transformation to an image range of image processing applications, including
will expand its low valued pixels to a higher level enhancement, analysis, restoration, and
and has little effect on higher valued pixels so in compression. The Fourier transform can also be
other words it enhances image in such a way that it used to perform correlation, which is closely
highlights minor details of an image Uses related to convolution. Correlation can be used to
locate features within an image; in this context
 Used to expand the values of dark pixels in an
correlation is often called template matching. The
image while compressing the higher values.
Fourier Transform is an important image
 It compresses the dynamic range of images with processing tool, which is used to decompose an
large variations in pixel values Proposed Design. image into its sine and cosine components. The
output of the transformation represents the image in
Propose a generalized equalization model for Fourier or frequency domain. In the Fourier
image & video enhancement; based on our analysis domain image each point represents a particular
on the relationships between image histogram and frequency contained in spatial domain image. The
contrast enhancement/ white balancing, we first FT is used in a wide range of application, such as
establish a generalized equalization model image analysis, image filtering, image
reconstruction and image compression. The FT is III. EXTENSION TO HISTOGRAM
used if we want to access the geometric EQUALIZATION
characteristics of a spatial domain image. It is easy
to examine or process certain frequency of the In 2006 to 2012 many works supported the
image in Fourier domain. In most implementations significance a contrast enhancement has an
Fourier image is shifted in such a way that the DC- important role in image processing applications.
value F (0,0) is displayed in centre of the image. They described that conventional contrast
The further away from the center an image point is enhancement technique either often fail to produce
the higher is its corresponding frequency. Bilog satisfactory results for a broad variety of low-
Transformation: Still there may be some presence contrast images. They described a new automatic
of negative frequency components (zero frequency method for contrast enhancement. First of all they
components). Bilog transformation is made use of grouped the histogram components of a low-
here to perform action on low frequency contrast image into a proper number of bins
information. The region near zeros is to be according to a selected criterion, then redistributed
highlighted for the enhancement and brightness these bins uniformly over the grayscale, and finally
preservation. Hence, after the application of this ungroup the previously grouped gray-levels.
transform, the region around zeros is enhanced. Accordingly, these new technique is named gray-
This is followed by grouping of pixels, where level grouping (GLG). GLG not only produces
clustering is done to increase the high resolution results superior to conventional contrast
pixels. At this stage, the image pixels are converted enhancement techniques, but is also fully automatic
back to RGB color model and pixels highlighted to in most incidents, and is suitable to a broad
a certain level Property: differences of images. An extension of GLG is
selective GLG (SGLG). SGLG selectively groups
 The bilog transform smoothly modifies the
and ungroup histogram components to achieve
gradient of the transformation so that in the region
specific application purposes. GLG was a general
near zero it remains finite. A single constant C is
and powerful technique, which can be suitable
provided to tune this behavior, so as to adjust the
manner, applied to a explicit variety of low-contrast
meaning of ―region near zero‖.
images and generates satisfactory results. HE
 The default value of this constant is 1/ ln(10); method could be conducted with full automation at
this gives a unity transfer function at zero but other fast speeds. In 2007 a modified Histogram
values can be applied as wished, to focus into the equalization (HE) has proved to be a simple and
region near zero or not. effective image contrast enhancement technique. It
worked on a novel technique called Multi-HE,
 The modified logarithmic transformation called which uniformly of decomposing the input image
Bilog transform can be both one-sided and into various sub-images, and then devoting the
symmetric, and thus can transform negative data to classical HE process to each one. This scheme
scaled negative data. performs a less increase produce image contrast
enhancement, in a way that the output image
 It can be applied to both the X and Y data, when presents a more natural look. It proposed two
it becomes a bi-symmetric log transform. Applying discrepancy functions for image decomposing,
log transformation to an image will expand its low imagining two new Multi-HE methods. A cost
valued pixels to a higher level and has little effect function was also used for automatically deciding
on higher valued pixels so in other words it in how many sub-images the input image will be
enhances image in such a way that it highlights decomposed on. The work was tested a new
minor details of an image. framework called MHE for image contrast
enhancement and brightness preserving which
NTSC: At the end process the NTSC technique is generated natural looking images. The results
operated on the pre-processed input image and showed that there methods was better on preserving
video to ensure the coloring effect on both. The the brightness of the processed image (in relation to
NTSC is especially for the video signal by means the original one) and yields images with natural
of using the YIQ model. appearance, at the cost of contrast enhancement.
Similarly init was stated that the HE technique was
not very well suited to be implemented in V. CONCLUSION
consumer electronics. They discussed that one of
the solutions to overcome this weakness is by In this paper, we have a tendency to analyze the
preserving the mean brightness of the input image relationships between image bar chart and
inside the output image. They provided the contrast/tone. We have a tendency to establish a
generalized effort model for world image tone
modified dualistic sub image HE method which
mapping. intensive experimental results
preserves the brightness of the image as shown in recommend that the projected technique has
Fig.2. They discussed results of first five methods sensible performances in several typical
that are available for contrast enhancement and applications together with image distinction
brightness preservation such as conventional global sweetening, tone correction, white reconciliation
HE, local HE, ADPHE, BBHE, DSIHE. The last and post-processing of de-hazed pictures. Within
method as MDSIHE gives better results than all the future, besides world image sweetening, we
have a tendency to expect to unify additional native
other. image sweetening strategies into the model through
native image feature analysis.
IV. EXPERIMENTAL RESULTS

Tone mapping for HDR Image is another natural V. REFERENCES

application for the proposed model. Many tone
mapping algorithms have been proposed through [1] T. Arici, S. Dikbas, and Y. Altunbasak, ―A
the years, e.g. those . Although the methods , based histogram modification framework and its
on local adaptive filtering achieve encouraging application for image contrast enhancement,‖ IEEE
results, the global method, such as gamma Trans. Image Process., vol. 18, no. 9, pp. 1921–
1935, 2009.
correction, is still the most popular choice because
[2] M. Ashikhmin, ―A tone mapping algorithm for
of its robustness and lower complexity. We test our high contrast images,‖ in Proc. 13th Eurographics
method on the HDR images captured by Nikon Workshop Rendering, 2002.
D7005 and map them into 8-bit and compare the [3] K. Barnard, L. Martin, B. Funt, and A. Coath,
results with those from the default tone mapping ―A data set for colour research,‖ Color Res.
process in MATLAB and gamma correction. Applicat., vol. 27, no. 3, pp. 147–151, 2002.
Although the default tone mapping in MATLAB [4] G. Buchsbaum, ―A spatial processor model for
object colour perception,‖ J. Frank. Inst., vol. 310,
can reveal some image details, it cannot recover the
1980.
color of image correctly. In other words, the [5] Z. Chen, B. Abidi, D. Page, and M. Abidi,
contrast is enhanced but the tone bias is raised. ―Gray-level grouping (glg): An automatic method
for optimized image contrast enhancement—Part i:
The basic method,‖ IEEE Trans. Image Process.,
vol. 15, no. 8, pp. 2303–2314, 2006.
[6] F. Ciurea and B. Funt, The SunBurst Resort,
Scottsdale, AZ, USA, ―A large image database for
color constancy research,‖ in Proc. IS&T/SIDs
Color Imaging Conf., 2004, pp. 160–164.
[7] D. Coltuc, P. Bolon, and J.-M. Chassery,
―Exact histogram specification,‖ IEEE Trans.
Image Process., vol. 15, no. 5, pp. 1143–1152,
2006.
[8] F. Drago, K. Myszkowski, T. Annen, and N.
Chiba, ―Adaptive logarithmic mapping for
displaying high contrast scenes,‖ in Proc. Computer
Graphics Forum, 2003, vol. 22, no. 3.
Fig.2. To each sub-figure, the left one shows [9] E. D. P. et al., ―Contrast limited adaptive
histogram image processing to improve the
original image which has obvious tonal
detection of simulated speculations in dense
distortion. The right one is the result gotten by mammograms,‖ J. Digital Imag., vol. 11, no. 4, pp.
the proposed method 193–200, 1998.
[10] J. K. et al., ―Deep photo: Model-based
photograph enhancement and viewing,‖ ACM
Trans. Graph.—Proc. ACM SIGGRAPH, vol. 27,
no. 5, 2008.
[11] Z. Farbman, R. Fattal, D. Lischinski, and R.
Szeliski, ―Edge-preserving decompositions for
multi-scale tone and detail manipulation,‖ ACM
Trans. Graph., vol. 27, no. 3, 2008.
[12] H. Farid, ―Blind inverse gamma correction,‖
IEEE Trans. Image Process., vol. 10, no. 10, pp.
1428–1433, 2001.
[13] R. Fattal, ―Single image dehazing,‖ ACM
Trans. Graph.—Proc. ACM SIGGRAPH, vol. 27,
no. 3, 2008.
[14] R. Fattal, M. Agrawala, and S. Rusinkiewicz,
―Multiscale shape and detail enhancement from
multi-light image collections,‖ ACM Trans. Graph.,
vol. 26, no. 3, pp. 51–59, 2007