6 abril 2011:Maquetación 1 11/08/2011 19:05 Página 87

Journal of Maritime Research, Vol. VIII. No. 1, pp. 87-100, 2011

Copyright © 2011. SEECMAR

Printed in Santander (Spain). All rights reserved

ISSN: 1697-4840

SIDESCAN SONAR IMAGERY PROCESSING

SOFTWARE FOR UNDERWATER RESEARCH AND
EDUCATION PURPOSES
M. I. Zamanillo1, J. M. Zamanillo2, E. Revestido3 and F. J. Velasco3
Received 11 February 2011; in revised form 16 February 2011; accepted 20 March 2011

ABSTRACT
Detailed submarine digital analysis of side scan sonar images significantly
enhances the ability to assess seafloor features and artifacts digital images. These
images are usually poor in their resolution if they are compared with optical
images. There are commercial solutions that could solve this trouble, such as: the
use of high resolution multibeam sidescan sonar, or the use of bathymetric sonar.
Present work shows an economical solution to avoid this kind of problem by using
digital image processing techniques under MATLAB environment. The applica-
tion presented here is easy to use and has been developed under user friendly phi-
losophy and could be operated for users at any level. Two types of sonar surveys,
seafloor mapping and submerged target searches (buried or not), each require dif-
ferent processing methods for data analysis. This work is the first step and a gener-
al purpose tool for future lines of research in submerged objects recognition.
Results are comparable in quality with commercial hardware solutions.

Keywords: SSS, single sidescan sonar, underwater, digital image processing,

autonomous underwater vehicle, remotely operated vehicle.

1 Professor, Electronics Technology, Systems and Automation Engineering Department (TEISA). University of Cantabria,
Higher Technical Navigation School, C/Gamazo nº 1 39004 Santander (Cantabria), Spain. Email: isabel.zamanilloB @unican.es,
Tel. +34 942 201331. 2 Professor, Communications Engineering Department (DICOM). University of Cantabria. I+D+i Telecom-
munication Building, Plaza de la Ciencia, Av. de los Castros s/n, 39005 Santander (Cantabria), Spain. Email: jose.zamanillo@uni-
can.es, Tel. +34943202219, Fax. +34942201488. 3 Professor, Universidad Electronics Technology, Systems and Automation En-
gineering Department (TEISA). University of Cantabria, Higher Technical Navigation School, C/Gamazo nº 1 39004 Santander
(Cantabria), Spain. Email: [email protected], Tel. +34942201368, Fax. +34942201303.

JOURNAL OF MARITIME RESEARCH 87

6 abril 2011:Maquetación 1 11/08/2011 19:05 Página 88

SIDESCAN SONAR IMAGERY PROCESSING SOFTWARE FOR UNDERWATER ...

INTRODUCTION
Digital images captured from the echoes of sidescan sonar onboard an unmanned
undersea vehicle, are usually characterized for their low resolution. This is so
because underwater, sound transmission is limited and this is most notable in use-
able ranges. The usable range of high frequency sound energy is greatly reduced by
seawater, typically to around 50 to 150 m (Blondel, P. 2009). Low frequency sound
energy is reduced at a much lesser rate with usable ranges of in excess of 250 m.
achievable. Therefore, a tradeoff exists between higher resolution images produced
by a high frequency side scan sonar and the longer range provided by a low frequen-
cy side scan sonar. In analyzing digital side scan sonar data numerous techniques
have been demonstrated to correct and enhance the imagery as well as aid in inter-
pretation (Lurton, X. 2002) (Medwin H. and Clay C.S. 1998)).
The collaboration of two research groups from different departments of the Uni-
versity of Cantabria have been necessary in the development of low cost and flexible
software for digital processing of sidescan sonar images under MATLAB environ-
ment by using digital filtering and advanced signal processing techniques. Some of
these techniques are already used in DVB digital video systems and broadcasting,
with excellent results (Hardie, R.C. and Barner, K.E. 1996).

MEASUREMENT TOOLS AND ENVIRONMENT

Advances in the fields of underwater technologies (Bellingham J.G. et al., 1994), robotics
(Ishoy A., 2000), acoustical positioning (Marani, G., 2009), (Palomeras N. et al., 2010), re-
mote sensing (Lillesand, Thomas M. and Kiefer, Ralph W. 1994), submarine guidance (An-
tich J., Ortiz A. and Oliver G., 2005) and digital processing imagery (Padmavathi, G.
et al.
2010) have led to the development of Autonomous Underwater Vehicles (AUVs) and
reached unimaginable levels only few years ago. Unlike other types of research can be car-
ried out inside a laboratory or in the field, measurement tools used in this work must be
adapted to a hostile environment as is the marine environment. These abilities have been
enabled better and precise researches in several fields, such as: biological (Dahms, Hans-
Uwe and Hwang, Jiang-Shiou, 2010), geological (Drury, Stephen A., 2001), zoological (Che-
ung, W.W.L. et al., 2009), natural resources (Wang Q. and Wang X., 2010), archeological
(Bowens A., 2009) and military (Von Alt C., et al., 2001). These characteristics have led to
the design of underwater vehicles with a built-in propulsion system, which has increased
the possibilities of movement in the underwater environment. This is the situation of UUVs
(Unmanned Underwater Vehicles) which are divided into two different categories:
— ROVs (Remotely Operated Vehicles), can be towed by a boat from the surface
at moderate speeds.
— AUVs (Autonomous Underwater Vehicles), which are propelled by electric
motors and governed autonomous remotely from the surface, from a ship or
from land.

88 Volume VIII. Number 1. year 2011

6 abril 2011:Maquetación 1 11/08/2011 19:05 Página 89

M. I. ZAMANILLO, J. M. ZAMANILLO, E. REVESTIDO AND F. J. VELASCO

With these systems, the researcher can move relatively quickly to a few meters
from the sea floor, following the topography of the same and recording video images
to find a place or object of interest. The problem presented by these vehicles is that
the image quality is degraded in terms of depth, and optical systems that capture
images need powerful lighting systems, which in the case of autonomous vehicles
(AUVs) without power connection through the umbilical wire to the surface is not
feasible and sidescan digital sonar images are used against optical ones. Figure 1 (a)
and 1 (b) show the ROV and AUV recently acquired by the University of Cantabria,
and used in this work. Figure 1 (a) shows the ROV, model Seaeye Falcon from the
Swedish company SAAB, this vehicle is an auxiliary rescue vehicle equipped with an
articulated arm. It is possible to use optical image recording underwater since it has
powerful illuminators fed from the surface. This vehicle lacks of a sonar device
(installation will be considered in the future) and this is the reason that direct com-
parison of optical images and the acoustic images is not available at this point.

Figure1: ROV and AUV owned by the University of Cantabria. (a) ROV, model Seaeye Falcon
from SAAB company. (b) AUV, model C’Inspector from Kongsberg company.

Figure 1 (b) shows the AUV, model C’Inspector from the Norwegian company
Kongsberg, this is an autonomous vehicle equipped with a high speed optical fiber data
connection of 1Km length. The vehicle can be used for inspection tasks in the back-
ground and detection of submerged objects and also it has been equipped with a.
Tritech SeaKing Sidescan Sonar with 675 kHz of operating frequency and chirp mod-
ulation. This device has a narrow beam and shorter range (100m) for more detailed
images of closer targets. Technical characteristics of this sonar are shown in Table I.
Chirp side scan sonar utilizes pulse compression techniques to produce long transmis-
sion pulses and achieve long range without a resultant decrease in across-track resolu-
tion. The commercial implementation of Chirp side scan sonar is in a single beam con-
figuration. Underwater, sound transmission is limited and this is most notable in use-
able ranges. The usable range of high frequency sound energy is greatly reduced by sea-
water, typically to around 50 to 100m (Blondel, P. 2009). Low frequency sound energy
is reduced at a much lesser rate with usable ranges of in excess of 200m achievable.

JOURNAL OF MARITIME RESEARCH 89

6 abril 2011:Maquetación 1 11/08/2011 19:05 Página 90

SIDESCAN SONAR IMAGERY PROCESSING SOFTWARE FOR UNDERWATER ...

Table I: Technical Characteristics From Tritech Seaking Sidescan Sonar.

Characteristics Value
Operating Frequency 675 kHz (chirp modulation)
Horizontal Beam width (-3dB) 1° 0.5°
Vertical Beam width (-3dB) 50°
Weight in air/water 5.3kg/2.7kg
Maximum operating depth 4000 m
Power Requirements 18-36V@12VA
Control Connector Tritech Sonar Connector
Transmitter Source Level 200 dB re 1μP @ 1 m
Transmitter Pulse Length 50 - 200 μs
Receiver Sensitivity > 2μV rms
Gain Control Range 80 dB
Display Dynamic Range 40 dB (Software Configurable)
Data Sampling Rates 5 - 200 μs
Data Resolution 4-8 bits (Software Configurable)
Software Tritech Seanet Display Software or low level language
commands
Data file format Proprietary Tritech “V4Log”
Communication Protocols Arcnet, RS-232
Communication data rates RS232 hasta 115.2 kbaud, Arcnet 156 ó 78 kbaud

Therefore, a tradeoff exists between higher resolution images produced by a high

frequency side scan sonar and the longer range provided by a low frequency side
scan sonar. Unfortunately, the AUV C’Inspector vehicle is not capable of providing
quality optical images below 1m in depth because the lack of external illuminators.
It is necessary to remark to refine and debug the software presented in this paper
authors have been used some images from the sidescan sonar of C’Inspector taken in
several measurement campaigns carried out in the Bay of Santander. But, most of the
sonar images have been used to test and refine the algorithms of digital processing
implemented in the software proceeds from free image libraries without copyrights
from different Internet sites.

SIDESCAN SONAR OPERATION AND ACOUSTIC SCATTERING THEORY

Single sidescan sonar devices transmit two beams, one on each side as it is shown
in Fig.2. These beams are narrow along-track to get a high resolution, and wide
across-track to cover as much range as possible. The distance from the sonar to a

90 Volume VIII. Number 1. year 2011

6 abril 2011:Maquetación 1 11/08/2011 19:05 Página 91

M. I. ZAMANILLO, J. M. ZAMANILLO, E. REVESTIDO AND F. J. VELASCO

point on the seabed or target is called the slant range. It should not be confused with
the ground range, between this point the point immediately below the sonar. The
angle of incidence of the incoming acoustic wave is a key factor in understanding
how it will scatter. Most of the energy is reflected in the specular direction. Some will
be reflected along other angles (scattering angles, distributed along the main refec-
tion angle). Depending on the seafloor or the submerged target, some energy will be
lost in the seabed. A very small portion, several orders of magnitude lower, might be
reflected back toward the imaging sonar also known as backscatter (Blondel and
Murton, 1997). The seabed reverberation area is mainly constituted by background
noises (Zhang Xiao-wei, Zheng Xiong-bo and Shen Yang, 2010).
The brightness of
Figure 2: Definitions of basic geometric parameters used in sidescan
sonar imagery.
sonar image is related
to the ratio between
the echoes to the
noise, if a comparison
with ordinary optical
images is made, sonar
images are low fre-
quency images and
they have less detail,
and the background
noises of sonar images
are high-frequency
impulse noises with
larger amplitudes rela-
tive to the multiple
echoes from the target
area. Because of the
Figure 3: Comparison between a sidescan sonar image and the photograph
of a tire in the Bay of Santander.
complexity of the un-
derwater environ-
ment, the gray level or
monochrome color of
sonar image from the
target area is usually
smaller than that of
(a) Sidescan sonar image with noise (b) Photograph image of the same tire made background noise. To
of a tire in the Bay of Santander. with the ROV improve the visual ef-
fects and reduce the
influences of the noise, it is very important to remove noises of sonar images, as it is
shown in Fig. 3 (a) for a grayscale sidescan image. Fig 3 (b) shows a photograph of the
same tire on a sandy seafloor.

JOURNAL OF MARITIME RESEARCH 91

6 abril 2011:Maquetación 1 11/08/2011 19:05 Página 92

SIDESCAN SONAR IMAGERY PROCESSING SOFTWARE FOR UNDERWATER ...

SIDESCAN SONAR ACQUISITION SOFTWARE

In present communication authors have been taking into account the geometrical
requirements shown in the prior section with the aim to get the best quality sonar
digital images. But, the sidescan sonar acquisition software has been provided by the
sonar manufacturer of the device and it has several limitations as shown below. The
standard file format use by the applications is the Tritech´s proprietary format
V4LOG which is not compatible with usual image processing tools that generally use
standard types of video and static image formats, to load the file such as: AVI (Audio
Video Interleave) or TIFF (Tagged Image File) formats for motion images and static
ones. Although, the application add the calibration grid, to the image file, which is a
serious limitation when the image is post-processed using digital techniques. This is
because the export filter makes a screen capture or rasterization (this actuates like a
flat bed scanner) of the image presented in the screen. To avoid this kind of problems
now we are working a new in-house control software using the low level commands
of the sidescan sonar.
During the first step of the research, our main effort has been focused on maxi-
mizing the quality of the original image provided by the commercial control soft-
ware, so we used a, uncompressed TIFF storage format. We do not use any other
most popular and compact graphic formats with uses compression such as JPG
(Joint Experts Group) to avoid the introduction of additional errors or compression
artifacts prior processing the image.

IMAGE EASY SONAR SOFTWARE

The interpretation of sonar images has traditionally been performed visually by
trained interpreters (Blondel, P. 2009); this feature presents the distinct advantage of
using the skill of the human interface to limits which are often unattainable by com-
puters. But there are also many disadvantages to a purely visual interpretation. First
of all, it is a subjective procedure: two interpreters with different experience, or dif-
ferent skills, are likely to get different interpretations for some features and details,
depending on their experience of the sonar used or of the environment studied.
Visual interpretation is also time-consuming, and a longer amount of time spent on
analysis does not ensure higher objectivity. As a result of this, present work has been
focused in the use of existing techniques applied in other areas of image digital pro-
cessing to the problem of sidescan sonar image processing. As a result of our
research, a simple and powerful software tool called ImageEasySonar, has been
developed. The application has been programmed under the friendly user philoso-
phy and could be operated by users at any level and would be used in educational and
training purposes. Although, the software could be running in different hardware
platforms such as: PC, Apple Macintosh, UNIX and, Linux machines because it has
been write under MATLAB environment. Another advantage consists in source code

92 Volume VIII. Number 1. year 2011

6 abril 2011:Maquetación 1 11/08/2011 19:05 Página 93

M. I. ZAMANILLO, J. M. ZAMANILLO, E. REVESTIDO AND F. J. VELASCO

can easily modified by the

programmer. In addition,
through integrated MAT-
LAB tool guide is possible
to develop professional
applications with elements
such as: pull-down menus,
pop-up menus, dialog
boxes, alerts, etc. without
the person responsible for
coding should make fur-
ther efforts to create such
objects. All these features,
combined with mathemat-
Figure 4: Aspect of the main interface of ImageEasySonar software.
ical optimization routines
and processing accompa-
nying the MATLAB software package has enabled that the first version of software
ImageEasySonar presented here has been operative in a relatively short time.

DIGITAL ALGORITHMS AND FILTERS USED IN IMAGE EASY

SONAR SOFTWARE
The spatial filtering operation techniques are applied to a digital image and highlight
or reduce details in order to simplify visual interpretation or provide further process-
ing. These techniques included in the image enhancement toolbox of MATLAB,
improve edge detail images and thus refocus digitally the scene digitally to reduce or
eliminate noise patterns in digital video before making a DVB transmission and they
are considered as local operations in digital image processing, in the sense of chang-
ing the value of each pixel in accordance with the values of the pixels that surround it,
transforming it comes to original levels so that they resemble or differ more than
those for neighboring pixels (Parker, J.R., 1996). The image enhancement algorithms
are applied to remotely sensed images in order to improve the appearance of an image
for the human visual analysis or occasionally for further computer analysis (Lee Y.
and Fam A., 1987). A possible classification of spatial filters based on its linearity,
being able to distinguish between linear and nonlinear filters. Within the first section
we can distinguish the spatial filters according to their spatial frequency. CS filters
(Comparison and Selection), (Lee Y. and Fam A, 1987), WMMR-MED (Weighted
Majority of M values with Minimum Range), (Longbotham, H. and Eberly, D., 1993),
Volterra (Taiho Koh and Powers, E., 1985) and EDGE (Canny, J. 1986), (Hardie, R.C,
Barner, K.E. 1996) , among others discussed in this paper correspond to the group of
non-linear enhancement filters. In addition to the aforementioned filtering tech-

JOURNAL OF MARITIME RESEARCH 93

6 abril 2011:Maquetación 1 11/08/2011 19:05 Página 94

SIDESCAN SONAR IMAGERY PROCESSING SOFTWARE FOR UNDERWATER ...

niques, it has been implemented in the program 14 additional filters from the image
processing MATLAB toolbox. The user of ImageEasySonar can select among 17 dif-
ferent image process techniques, which are widely detailed in the literature (Canny, J.
1986), (Corinthios, M., 1999), (Longbotham, H. and Eberly, D., 1993), (Mitra, S.K., et
al., 1991), (Jensen, John R. 1996), (Zhuo, S. Guo D. and Sim, T., 2010).
The Comparison and Selection (CS) filter is one of the simpler enhancement filters
(Lee Y. and Fam A, 1987). As an example, we give a brief explanation about this filters
works. The first step is to choose the color space to apply the transformation, if grayscale
is chosen the mathematical transformation is apply to one layer. If RGB space is chosen
it is necessary to apply separately the technique layer by layer (Red layer, Green layer
and Blue layer) as it is shown in figure 5(a) and Fig 5 (b). Values are in the integer range
from 0 up to 255. This feature expands by other tree the computation time.
The second step consist in
properly choose the geometry
(linear, square, circular, etc)
and size (3x3 pixels, 5x5 pixel,
9x9 pixels or 11x11 pixels), of
(a) B&W photograph
the movable exploring window
of the lighthouse of or kernel, which size makes
Mouro located in de
Bay of Santander (one
more visible the edges as
layer is only needed to shown in Fig. 6 (c), or soften as
process the image)
it is shown in figure 6 (b).
Third step consist to properly
fix the J parameter of the algo-
rithm, this parameters fixed
the distance (in pixels) from
the mean of all numerical val-
ues of the kernel renumbered
from reordered from the min-
imum to the maximum ac-
cording with the following.
(b) Color photograph of the The output Yk of the CS
lighthouse of Mouro located
in de Bay of Santander (three filter with parameter J at time
layer R, G and B are needed to k is defined through the input
process the image)
values Xk-N, …, Xk+ N, in a
Figure 5: The color space is an important election to apply spatial window of length 2 N+ 1, for
digital filtering. a positive integer N by:

¬¯ ( N 1 J ), if Nk s M k
Yk  ­ X k( N 1 J ) (1)
¯® X k , otherwise

94 Volume VIII. Number 1. year 2011

6 abril 2011:Maquetación 1 11/08/2011 19:05 Página 95

M. I. ZAMANILLO, J. M. ZAMANILLO, E. REVESTIDO AND F. J. VELASCO

where Xk(i) is the ith smallest sample inside the window, μk and Mk are the sample
mean and median, respectively, and J is an integer satisfying 1 ≤ J ≤ N.
When J = N, the CS filter selects either the minimum or the maximum value in
the window depending on the data. For the case of J = 0, this filter reduces to the well
investigated median filter. Since CS filters are nonlinear, the superposition property
does not hold in its general form. (Lee Y. and Fam A., 1987). The CS filter ranks the
values in the filter window or kernel in numerical order, and calculates the mean
value. Parameter J identifies a pair of rank numbers (measured inward from the top
and bottom of the rank list) whose corresponding raster values provide the two pos-
sible filter output values. If the center cell value is less than the window mean, the
lower output value is assigned, and if it is greater than the mean, the higher output
value is used. The CS filter sharpens blurred edges while smoothing non-edge areas.
The sharpening effect increases with lower values of Parameter J (which move the
filter output values farther from the mean).
Some of the results obtained when applying digital filtering techniques just been
discussed and depicted in Figure 6. For brevity the mathematical expressions used in
each filtering technique are not shown here and we have ignored many of the results
obtained for other images. Another interesting feature of the application is that user
can apply several different digital filters on the same image and visualize graphically,
original image and up to two different transformations before save them into the
hard disk An important parameter the execution time of each digital filtering
process, which is dependent on the selected parameters in the application as shown
in Figure 6 (b) and Figure 6 (c) obtained as the digital processing from the image
shown in Figure 6 (a).
Figure 6 (d) shows the result of applying the algorithm edge (it is noticeable that
the calibration grid is visible) to the image, while filtering when combined CS plus
edge filtering, result not only removes the grid also shows in detail the edges of the
processed image. The edge detection class is designed to detect and highlight bound-
aries between image areas that have distinctly different brightness. The output raster
is a grayscale image of the edges, with the cell brightness proportional to the differ-
ence in neighboring cell brightness in the original image. The resulting image can be
used as the basis for additional image interpretation and analysis, such as image seg-
mentation.
Filtering techniques based on the Prewitt filters and/or Sobel shown in Figure 6
(f) and Figure 6 (g) require a lower computational cost than the above-mentioned
CS, and EDGE WWMED. In our case, the result not only gives different color to the
original image but it gives an appearance of high-relief, or false 3D. Results are simi-
lar results to those obtained with more expensive hardware. As for future work,
authors wish to apply these techniques in real-time time or at least “quasi-real” time
in this part of the research the main goal has been the computational speed versus
quality processing.

JOURNAL OF MARITIME RESEARCH 95

6 abril 2011:Maquetación 1 11/08/2011 19:05 Página 96

SIDESCAN SONAR IMAGERY PROCESSING SOFTWARE FOR UNDERWATER ...

(a) Original Image (with grid) (b) Processed image with a CS Filter (3x3, J=4 window),
(no grid is visible)

(c) Processed image with a CS Filter (7x7, J=2 (d) Processed image with an EDGE filter
window), (grid appears again) (grids are also detected)

(e) Processed image with a CS 3x3 y J=0.4 (f) Result of apply a Prewitt filter with the original image
followed by an EDGE filter

(g) Result of apply a Sobel filter with the original image

(h) Aspect of the Zoom tool of ImageEasySonar

Figure 6: Aspect of the interface and results of digital processing sidescan sonar images
using ImageEasySonar softwar.

96 Volume VIII. Number 1. year 2011

6 abril 2011:Maquetación 1 11/08/2011 19:05 Página 97

M. I. ZAMANILLO, J. M. ZAMANILLO, E. REVESTIDO AND F. J. VELASCO

The hardware and software characteristics used in this study, are presented here:
— PC compatible with μP AMD Athlon64 X2 Dual Core @ 3.01 GHz y 2 GB
RAM DDR memory.
— Operative System Windows XP Professional v2002 with Service Pack2.
— MATLAB R2009b 7.9.0.
CPU computation time is less than 5% with about 40 processes running at that
time and increasing from 50 to 55% during the processing of the filter applied. The
memory usage during the process of computing including loading the program itself
in any case not exceeds 90 MBytes. Figure 6 (h) shows the zoom tool included in the
application that allows seeing the details more clearly, even compared to the initial
image. The novelty of the work presented here is the use of these techniques
designed for static and DVB video optical captured images to “acoustically” ones
from low resolution single sidescan sonar.

CONCLUSIONS
It has developed an application software that allows processing images from a sides-
can sonar by using techniques of post-digital signal processing, some of them
already used in DVB digital video systems. The intuitive interface to use has been
programmed under user-friendly, and WYSIWYG (What you see is what you get)
philosophies and could be operated for users at any level for research or educational
purposes. Furthermore, the application allows the overlapping of different filtering
techniques and improving the image quality on the same box or frame. By applying
the Prewitt and Sobel filters is possible to get an approximate high relief profile of the
seafloor without the need of acquires expensive hardware like: multibeam side-scan
sonar or a slow scan by bathymetric sonar.

ACKNOWLEDGMENTS
Authors would like to thank the Scientific and Technical Services Research (SCTI),
of the University of Cantabria under the supervision of Vicerrectorado de Investi-
gación y Transferencia Tecnológica, and the Regional Government of Cantabria for
the logistical and financial support without this work has not been realized.

JOURNAL OF MARITIME RESEARCH 97

6 abril 2011:Maquetación 1 11/08/2011 19:05 Página 98

SIDESCAN SONAR IMAGERY PROCESSING SOFTWARE FOR UNDERWATER ...

REFERENCES

Antich, J. et al. (2005). "A PFM-based control architecture for a visually guided underwater
cable tracker to achieve navigation in troublesome scenarios". Journal of Maritime
Research, Vol. II, No. 1, pp. 33-50.
Bellingham, J.G. et al. (1994): A second generation survey AUV, in Proc. Autonomous Under-
water Vehicle Technology, July pp. 148–155.
Blondel, P. (2009): The Handbook of Sidescan Sonar. Springer/Praxis, Heidelberg, Ger-
many/Chichester, U.K., 316 pp.
Blondel, P. and B.J. Murton (1997): Handbook of Seafloor Sonar Imagery. Wiley/Praxis,
Chichester, U.K.
Bowens, A. (2009): Underwater Archaeology: The NAS Guide to Principles and Practice Sec-
ond Edition, Blackwell Publishing, UK, 220pp.
Canny, J. (1986): A Computational Approach to Edge Detection, IEEE Transactions on Pat-
tern Analysis and Machine Intelligence, Vol. PAMI-8, No. 6, pp. 679-698.
Cheung, W.W.L.; Lam, V.W.Y.; Sarmiento, J.L.; Kearney, K.; Watson, R. and Pauly, D. (2009):
Projecting global marine biodiversity impacts under climate change scenarios. Fish and
Fisheries, Willey. Vol. 10, pp. 235–251.
Corinthios, M. (1999): Signals, Systems, Transforms, and Digital Signal Processing with
MATLAB. CRC Press, pp. 1316.
Dahms, Hans-Uwe and Hwang, Jiang-Shiou, (2010) :Perspectives of underwater optics in bio-
logical oceanography and plankton ecology studies. Journal of Marine Science and Tech-
nology, Vol. 18, No. 1, pp. 112-121
Drury, Stephen A. (2001): Image Interpretation in Geology (3rd ed.). Chapter 5, Digital Image
Processing. New York: Routledge. pp. 133-138.
Hardie, R.C.; Barner, K.E. (1996): Extended permutation filters and their application to edge
enhancement. IEEE Transactions on Image Processing, Vol. 5, No. 6, Jun, pp.855-867
Hardie, R.C. and Boncelet, C., (1993): LUM filters: a class of rank-order-based filters for
smoothing and sharpening, IEEE Transactions on Signal Processing, Vol. 41, pp. 1061-
1076.
Hardie, R.C. and Boncelet, C.G., (1995): Gradient-based edge detection using nonlinear edge
enhancing prefilters, IEEE Transactions on Image Processing, Vol.4 (11), 1572-1577.
Ishoy, A. (2000): How to make survey instruments “AUV-friendly”,” in Proc. MTS/IEEE Int.
Conf. OCEANS, pp. 1647–1652.
Jensen, J. R., (2005): An Introductory Digital Image processing: A remote sensing perspective
Image enhancement, Prentice-Hall, UpperSaddle River N.J. 526 pp.
Jensen, John R. (1996): Introductory Digital Image Processing: a Remote Sensing Perspective
(2nd ed.). Chapter 7, Image Enhancement. Upper Saddle River, NJ: Prentice-Hall. pp.
153-172.

98 Volume VIII. Number 1. year 2011

6 abril 2011:Maquetación 1 11/08/2011 19:05 Página 99

M. I. ZAMANILLO, J. M. ZAMANILLO, E. REVESTIDO AND F. J. VELASCO

Lee Y. and Fam A, (1987): An edge gradient enhancing adaptive order statistic filter. IEEE
Transactions on Acoustics, Speech, and Signal Processing Vol.35, pp. 680-695
Lillesand, Thomas M. and Kiefer, Ralph W. (1994): Remote Sensing and Image Interpretation
(3rd ed.). Chapter 7, Digital Image Processing. New York: John Wiley and Sons. pp. 585-
618.
Lim, Jae S.,(1990): Two-Dimensional Signal and Image Processing, Englewood Cliffs, NJ,
Prentice Hall, pp. 478-488.
Longbotham, H. and Eberly, D., (1993): The WMMR filters: a class of robust edge enhancers, ,
IEEE Transactions on Signal Processing Vol. 41, pp. 1680-1685.
Lurton, X. (2002): An Introduction to Underwater Acoustics. Springer/Praxis, Heidelberg,
Germany/Chichester, U.K., 380 pp.
Marani, G. (2009): Advances in autonomous underwater intervention for AUVs, in Work
Shops and Tutorials, 2009 IEEE International Conference on Robotics and Automation.
Kobe, Japan.
Medwin, H. and Clay, C.S. (1998): Fundamentals of Acoustical Oceanography. Academic
Press, London, 712 pp.
Mitra, S.K.; Li, H.; Lin, I.-S. and Yu, T.-H. (1991): A new class of nonlinear filters for image
enhancement, Acoustics, Speech, and Signal Processing, 1991. ICASSP-91., 1991 Interna-
tional Conference on 4, 2525-2528.
Padmavathi, G.et al. (2010): Comparison of Filters used for Underwater Image Pre-Process-
ing. IJCSNS International Journal of Computer Science and Network Security, VOL.10
No.1, January, pp.58-65.
Palomeras, N. et al.,(2010): Distributed Architecture for Enabling Autonomous Underwater
Intervention Missions, in Proceedings Of the 4th Annual IEEE International System
Conference. San Diego, USA.
Parker, James R., (1996): Algorithms for Image Processing and Computer Vision, New York,
John Wiley & Sons, Inc., pp. 23-29.
Poularikas, A.D. and Ramadan, Z.M. (2006): Adaptative Filtering Primer with MATLAB. Tay-
lor & Francis. CRC Press.
Russ, J.C. (2007): The Image Processing Handbook. 5th Edition. CRC Press.pp.817.
Taiho Koh and Powers, E., (1985): Second-order Volterra filtering and its application to non-
linear system identification, IEEE Transactions on Acoustics, Speech, and Signal Process-
ing, Vol. 33, pp. 1445-1455.
Thurnhofer, S. and Mitra, S.K., (1996): A general framework for quadratic Volterra filters for
edge enhancement, IEEE Transactions on Image Processing, Vol. 5 pp. 950-96.
Urick, R.J. (1996): Principles of Underwater Sound, Third Edition. Peninsula Publishing, Los
Altos, CA, 444 pp.
Uvais Qidwai and C.H. Chen: Digital image processing, An algorithmic approach with MAT-
LAB. Chapman and Hall Textbooks in computing. CRC Press ISBN 978-1-4200-7950-0.

JOURNAL OF MARITIME RESEARCH 99

6 abril 2011:Maquetación 1 11/08/2011 19:05 Página 100

SIDESCAN SONAR IMAGERY PROCESSING SOFTWARE FOR UNDERWATER ...

Von Alt, C., et al. (2001): Hunting for mines with REMUS: A high performance, affordable,
free swimming underwater robot, in Proc. MTS/IEEE Int. Conf. OCEANS, pp. 117-122.
Wang, Q. and Wang, X. (2010): Interferometric Fiber Optic Signal Processing Based on
Wavelet Transform for Subsea Gas Pipeline Leakage Inspection. International Confer-
ence on Measuring Technology and Mechatronics Automation (ICMTMA 2010), pp.
501-504.
Zhang Xiao-wei; Zheng Xiong-bo and Shen Yang, (2010): Sidescan sonar image de-noising
algorithm in multi-wavelet domain. Computer Application and System International
Conference on Modeling (ICCASM 2010), Vol: 2, pp. 367-371.
Zhuo, S.; Guo D. and Sim, T. (2010): T. Robust Flash Deblurring, IEEE Conference on Com-
puter Vision and Pattern Recognition (CVPR), pp. 2440-2447.

100 Volume VIII. Number 1. year 2011