Visual Comparison of Underwater Image

Enhancement Algorithms

Sarvesh Milind Mohite Shubham Pramod Pawar

Dept. Electronics Engineering Dept. Electronics Engineering
Ramrao Adik Institute Of Technology,Sector 7, Ramrao Adik Institute Of Technology,Sector 7,
Nerul, Navi Mumbai,400706. Nerul, Navi Mumbai,400706.
[email protected] [email protected]

Mohit Milind Shukla Smitha Raveendran

Dept. Electronics Engineering Dept. Electronics Engineering
Ramrao Adik Institute Of Technology,Sector 7, Ramrao Adik Institute Of Technology,Sector 7,
Nerul, Navi Mumbai,400706. Nerul,Navi Mumbai,400706.
[email protected] [email protected]

Abstract— The essential target of underwater image deeper but also colors drop off one by one depending on the
enhancement is to enhance the quality of a distorted image by wavelength of the colors.
keeping the intrinsic information of the image preserved. The
main difficulty in image enhancement is determining the II. INFRASTRUCTURE AND IMPLEMENTATION
criteria for enhancement and, therefore, more than one image
enhancement techniques are empirical and require interactive
procedures to obtain satisfactory results. In this report robust
image enhancement algorithms are discussed, implemented to
distorted images and compared according to their robustness.
The algorithms are especially able to improve the contrast of
underwater images by means of software techniques. When
deciding that one image has better quality than another image,
quality measure metrics are needed. Otherwise comparing
image quality just by visual appearance may not be objective
because images could vary from person to person. That is why
quantitative metrics are crucial to compare images for their
qualities.

I. INTRODUCTION
Underwater image enhancement techniques provide a way Fig 1. Setup
to improving the object identification in
underwaterenvironment. There is lot of research started for ABOUT THE SETUP :
the improvement of image quality, but limited work has
This setup contains mainly four components, analog audio,
been done in the area of underwater images, because in
master-node architecture, sensors and measurement device
underwater environment image get blurred due to poor
and camera and image processing. The scope of this system
visibility conditions and effects like absorption of light,
is to capture underwater images, detect objects, acoustic
reflection of light, bending of light, denser medium (800
noise jamming. The above Fig 1 is a setup which is located
times denser than air), and scattering of light etc. These are
at RAIT college laboratory. For our research based project,
the important factor which causes the degradation of
the images which was captured is the college logo and the
underwater images [1]. Another well-known problem
distorted image was enhanced using various filtering
concerning the underwater images is related to the
techniques and the results and conclusions are presented
density of the water in the sea which is considered 800 times
below. Below fig 2. Represents the setup components block:
denser medium than air. Therefore, when light rays moves
from the air to the water, it is partly reflected reverse and at
the same time partly enters the water. The total light amount
that enters the water also starts reducing start it goes deeper
in the sea. Similarly, the water molecules also absorb certain
amount of light. As a result, the underwater images are
getting darker and darker as the deepness increases. Not
only the quantity of light rays is condensed when it goes
 III. DESCRIPTION limit the height of the histogram to a certain level we can
limit the slope of the CDF and hence the amount of contrast
A. Adaptive Histogram Equalization enhancement. The only difference between regular AHE and
Adaptive histogram equalization (AHE) is a computer CLAHE is that there is one extra step to clip the histogram
image processing technique used to improve contrast in before the computation of its CDF as the mapping function
images. It differs from ordinary histogram equalization in is performed. Hence CLAHE is implemented in the same
the respect that the adaptive method computes several function tiledAHE in ahe.cpp. The program ”ahe” takes an
histograms, each corresponding to a distinct section of the additional optional parameter which specifies the level at
image, and uses them to redistribute the lightness values of which to clip the histogram. By default no clipping is
the image. It is therefore suitable for improving the local performed. Valid values for clipping fall in the range from 1
contrast and enhancing the definitions of edges in each to 1/bins.
region of an image. However, AHE has a tendency to
overamplify noise in relatively homogeneous regions of an
Images Used for Filtering Process:
image. A variant of adaptive histogram equalization called
contrast limited adaptive histogram equalization (CLAHE)
prevents this by limiting the amplification. Ordinary
histogram equalization uses the same transformation derived
from the image histogram to transform all pixels. This
works well when the distribution of pixel values is similar
throughout the image. Adaptive histogram equalization
(AHE) improves on this by transforming each pixel
with a transformation function derived from a
neighbourhood region. It was first developed for use in
aircraft cockpit displays. In its simplest form, each pixel is
transformed based on the histogram of a square surrounding
the pixel, as in the figure below. The derivation of the
transformation functions from the histograms is exactly the
same as for ordinary histogram equalization: The
transformation function is proportional to the cumulative
distribution function (CDF) of pixel values in the
neighbourhood.
B. Contast Stretching Fig 2. RAIT Logo
Contrast stretching (also called Normalization) attempts to
improve an image by stretching the range of intensity values
it contains to make full use of possible values. Unlike
histogram equalization, contrast stretching is restricted to a
linear mapping of input to output values. The result is less
dramatic, but tends to avoid the sometimes artificial
appearance of equalized images.The problem with this is
that a single outlying pixel with either a very high or very
low value can severely affect the value of c or d and this
could lead to very unrepresentative scaling. Therefore a
more robust approach is to first take a histogram of the
image, and then select c and d at, say, the 5th and 95th
percentile in the histogram (that is, 5% of the pixel in the
histogram will have values lower than c, and 5% of the
pixels will have values higher than d). This prevents outliers
affecting the scaling so much.
C. CLAHE
While performing AHE if the region being processed has a
relatively small intensity range then the noise in that region
gets more enhanced. It can also cause some kind of artifacts
to appear on those regions. To limit the appearance of such Fig 3. Database Image
artifacts and noise, a modification of AHE called Contrast
Limited AHE can be used. The amount of contrast
enhancement for some intensity is directly proportional to
the slope of the CDF function at that intensity level. Hence
contrast enhancement can be limited by limiting the slope of
the CDF. The slope of CDF at a bin location is determined
by the height of the histogram for that bin. Therefore if we
Implementation of Enhancement Techinques on Implementation of Enhancement Techinques on
Image taken from the setup Database Image

Fig 4. Original Image Fig 11. Original Image

Fig 5. AHE Fig 6. Histogram for AHE Fig 12. AHE Fig 13. Histogram: AHE

Fig 7. CLAHE Fig 8. Histogram for CLAHE Fig 14. CLAHE Fig 15. Histogram:CLAHE

Fig 9. CS Fig 10. Histogram for CS Fig 16. CS Fig 17. Histogram: CS
Results obtained after implementation of [7] Iqbal K, Odetayo M, James A, Salam R A and
enhancement techniques: Talib A 2010 Enhancing the low quality images using
1. Implementing the adaptive histogram equalization
unsupervised colour correction method. IEEE Int.
method, gained us the image where small details Conf. Syst. Man Cybern. (SMC), pp 1703–1709
such as letters where easily visible.
[8] Kwok N, Wang D, Jia X, Chen S, Fang G and Ha
2. Implementing the CLAHE method, gave us the Q 2011 Gray world based color correction and
overall crisp image and the distortion was reduced intensity preservation for image enhancement. Int.
with respect to the original image. Congress Image Signal Process. (CISP) 2: 994–998
3. Implementing the contrast stretching, gained us the [9] Provenzi E, Gatta C, Fierro M and Rizzi A 2008 A
result where contrast of the image was decreased spatially variant white-patch and gray-world method
and the image was more clear if there were low
for color image enhancement driven by local contrast.
light conditions.
IEEE Trans. Pattern Anal. Mach. Intell. 30(10): 1757–
1770.
V. CONCLUSION

Enhancing the Underwater image through various

filtering techniques gave us a analysis of effects of
filters on the Underwater image. Three different
methods were being used in this entire process.Thus,
this helped us to enhance the quality of underwater
image to a greater extent. From the above impletation
of various filter we found out that CLAHE filtering
technique gives us the better image quality.

VI. REFERENCES

[1] Chiang J Y and Chen Ying-Ching 2012

Underwater image enhancement by wavelength
compensation and dehazing. IEEE Trans. Image
Process. 21(4): 1756–1769

[2] Pegau W S, Gray D and Zaneveld J R V 1997

Absorption and attenuation of visible and near-infrared
light in water: Dependence on temperature and
salinity. Appl. Opt. 36(24): 6035– 6046

[3] Schechner Y Y and Karpel N 2005 Recovery of

underwater visibility and structure by polarization
analysis. IEEE J. Oceanic Eng. 30(3): 570–587

[4] Sedlazeck A and Koch R 2011 Simulating deep sea

underwater images using physical models for light
attenuation, scattering, and refraction. Vision,
Modeling, and Visualization Workshop, pp 49–56

[5] Trucco E and Olmos-Antillon A T 2006 Self-

tuning underwater image restoration. IEEE J. Oceanic
Eng. 31(2): 511–519

[6] Iqbal K, Salam R A, Osman A and Talib AZ 2007

Underwater image enhancement using an integrated
colour model. IAENG Int. J. Comput. Sci. 32(2): 239–
244