Ocean Engineering 117 (2016) 39–44

Contents lists available at ScienceDirect

Ocean Engineering
journal homepage: www.elsevier.com/locate/oceaneng

Deflection measuring system for floating dry docks

Valery V. Korotaev a, Anton V. Pantiushin a,n, Mariya G. Serikova a, Andrei G. Anisimov b
a
Department of Optical-Electronic Devices and Systems, ITMO University, 197101, Saint Petersburg, Kronverksky 49, Russia
b
Aerospace Non-Destructive Testing Laboratory, Faculty of Aerospace Engineering, Delft University of Technology, Postbus 5058, 2600 GB Delft,
The Netherlands

art ic l e i nf o a b s t r a c t

Article history: Reliable measurement of a deflection magnitude refers to one of the most important tasks in floating dry
Received 25 June 2015 dock operation. Current and accurate information about the deflection can prevent structural damage
Accepted 6 March 2016 and failure by activating alarms and on-board deflection compensation systems. This paper describes the
Available online 22 March 2016
development of a deflection measuring system for floating dry docks. The measuring system is capable of
Keywords: measuring the actual deflection of a dock structure in real-time and fully automatic mode. The system
Deflection measurement consists of a set of reference marks based on light emitting diodes and a measurement unit with two
Floating dry dock oppositely directed camera-based channels. The measurement unit is placed in the middle of a wing deck
Camera-based measuring system of the dock. The measurement unit and the reference marks are aligned along the wing deck. The
Reference mark
accuracy of deflection measurement of 7 1:5 mm in the range of 7150 mm for the docks in length more
Harsh environment
than 100 m was achieved during laboratory tests. The system resistance to harsh environment (salt fog,
temperature gradient, high humidity, etc.) and the ability to perform measurements in 24/7 regime was
proved through field tests at a floating dry dock in operation.
& 2016 Elsevier Ltd. All rights reserved.

1. Introduction effective deflection compensation. At the same time current

Floating dry docks are large scale structures designed for bui-
lding and maintaining ships. Dimensions of floating dry docks
(length, width, and depth) depend on the size of ships to be
docked. As sizes of modern ships tend to grow, the length of the
modern docks can exceed 200 m, while the height and width can
reach 50 and 100 m, respectively (Gaythwaite, 2004). As it is a
huge structure, a dry floating dock experiences heavy loads both
from its own weight and from a docked ship. These loads can
significantly vary in time during floating or refloating stage of
docking procedure, and also depend on the type of the docked
ship and its location on the pontoon deck. Moreover, the loads
usually have a highly uneven distribution owing to the complexity
of the weight distribution of the docked ship, as shown in Fig. 1.
This leads to uneven deformations of the dock structure. Floating
dry docks mainly suffer from longitudinal deflection (plane Y 0 Z 0 in
Fig. 1), rather than in transverse directions (plane X 0 Y 0 ). Excessive
allowable deflections can cause severe structural damage and
failure. Floating dry docks typically have a system of ballast tanks
to compensate the arising deformations. But this system requires
accurate and reliable information about the current dock deflec-
tion distribution as an input parameter in order to provide
n
Corresponding author. Tel.: þ 7 812 595 41 59.
E-mail address: pantiushin@niuitmo.ru (A.V. Pantiushin).
deflection is of general interest for the dock crew. That is why the
deflection measurement is one of the most important tasks in
floating dry dock operation.
Several types of deflection measurement systems for floating dry
docks were reported, among them: strain measurement devices
based on strain gauges or optical fiber sensors (Froggatt and Moore,
1998; Sannerhaugen and Hellvik, 1999; Zou et al., 2006), deflection
and inclination measurement systems based on wired and wireless
sensor networks (Lynch et al., 2006; Yang et al., 2013), inspection
based on geodetic devices (Stiros and Psimoulis, 2012; Carbonari
et al., 2013) and automatic camera-based measurement systems
(Newman and Jain, 1995; Korotaev et al., 2000; Gorbachev et al.,
2007). Besides of strong points, each type of systems has limita-
tions. Therefore, according to international regulations in safety a
dock has to be equipped with at least two deflection measurement
systems based on different physical measuring principles (Germa-
nischer Lloyd Aktiengesellschaft. Rules for Classification and Con-
struction. Floating Docks, 1993; Rules for Classification of Floating
Docks, 2009; Rules for Building and Classing. Steel Floating Dry
Docks, 2009; Rules for Technical Supervision during Construction of
Ships and Manufacture of Materials and Products for Ships, Part V
Technical Supervision During Construction of Ships, 2014).
Strain measurement systems allow measurement of local strain
changes of an optical fiber distributed along the hull structure
(Froggatt and Moore, 1998; Sannerhaugen and Hellvik, 1999; Zou
et al., 2006). A central unit calculates strain profile based on

http://dx.doi.org/10.1016/j.oceaneng.2016.03.012
0029-8018/& 2016 Elsevier Ltd. All rights reserved.
40 V.V. Korotaev et al. / Ocean Engineering 117 (2016) 39–44

Fig. 2. An approximation of the dock wing wall deformation by polynomial.

The defined measurement task can be resolved by an imple-

Fig. 1. A floating dry dock and a load distribution of a typical ship under inspection. mentation of a measuring system, which is capable of measuring a
spatial position of predefined points along the wing deck of the
Table 1
dock. Special markers or sensors could be placed in these pre-
Measurement task specification. defined points in order to visualize their displacements or to
measure the displacements directly. The points location along the
Parameter Value deck have to be defined by dock designer, however, points with
the highest displacements are of interest.
Dock length up to 200 m
Measurement range along Y 0 -axis vertical direction) 7 100 mm According the material strength theory (Smith and Hashemi,
Measurement accuracy along Y 0 -axis 7 5 mm 2006; Mott, 2014) and practice, the magnitude of the dock
Measurement resolution 0.1 mm deflection depends on the distance to the power of four, so the
dock deflection profile can be approximated by fourth-degree
polynomial. This polynomial determines the deformation of the
measured data. One of the difficulties of this method imple-
construction at distance z from the origin of the coordinate system
mentation is a complexity of fiber sensor installation along the
presented in Fig. 2. So the measurement procedure for reliable
whole hull with multiple bays and bulkheads. The inclination- estimation of the actual deformation profile of the floating dock
based method also suffer this inconvenience (Vincent et al., 2008; requires the next steps:
Swartz et al., 2012). Deflection calculation in this case is based on a
relative angle displacement of sensors installed along the hull Step 1. A prior measurement of the initial deformation profile
(Lynch et al., 2006; Yang et al., 2013). The angular displacement hinitial(z) at the unloaded state.
can be measured by accelerometers, liquid inclinometers and Step 2. Measurement of the deformation profile h(z) at the
other techniques (Smith and LeVezu, 2012). loaded state.
Optical and camera-based measurement methods define a Step 3. Measurement of the normalized deformation profile at
reference line, a so-called sighting line, so the transverse dis- the loaded state hnormalized ¼ hðzÞ  hinitial ðzÞ.
placements of an observed point are measured in the coordinate
system based on this line (Maraev and Timofeev, 2013; Maraev et When the displacements hðz1 Þ; hðz2 Þ; …; hðzn Þ at the chosen
al., 2014; Kleshchenok et al., 2014). According to the practice points with positions z1 ; z2 ; …; zn are measured, the deformation
manual vision inspection performed by a person with conven- profile of the wing deck can be calculated by solving the following
tional optical measurement systems is time consuming and system of equations:
includes human factor (Korotaev et al., 2012). Modern camera- 2 3
2 4 3 k4 2 3
based solutions have the ability to fully automate the process due z1 z31 z21 z1 1 6 7 hðz1 Þ
6 4 k
76 3 7 6 7
to recent advances in machine vision techniques by automatic 6 z2 z32 z22 z2 1 76 7 6 hðz2 Þ 7
6 76 7 6 7
target detection and data processing (Newman and Jain, 1995; 6 ⋮ ⋮ ⋮ ⋮ ⋮ 76 k 2 7 ¼ 6 ⋮ 7: ð1Þ
4 56 k 7 4 5
Nixon et al., 2012; Sonka et al., 2014). 4 1 5
z4n z3n z2n zn 1 hðzn Þ
This paper introduces an approach to deflection measuring of a k0
floating dry dock. The approach is based on a non-contact mea- where k4 ; k3 ; k2 ; k1 ; k0 are the polynomial coefficients to be ide-
suring of linear shifts of finite number of marks placed along a ntified.
wing wall of a dock. The measurements are done by a designed In most of the cases (especially for a 100 m dock or less) the
and developed camera-based system that employs machine-vision deflection of the dock represents a parabola (Hedger, 2007). Thus,
techniques. A camera-based solution was chosen for a non-contact the minimum number of the measurement points can be reduced
method of measuring which minimizes a number of sensors and to three, because only three unknowns k2 ; k1 ; k0 have to be defined
wires to be installed at a dock. The camera-based solution also has to fully describe second-degree polynomial, which is a parabola. In
a high potential for an automation and has a low dependence on this case, two measurement points have to be located at each end
the human factor. of the wind deck, third one has to be placed in the middle of the
deck. Then, Eq. (1) can be simplified as follows:
2 2 32 3 2 3
z1 z1 1 k2 hðz1 Þ
2. Deflection measuring of floating dry docks: Task definition 6 z2 z 7
1 54 k1 5 ¼ 4 hðz2 Þ 7
6 7 6
4 2 2 5: ð2Þ
z23 z3 1 k0 hðz3 Þ
Deflection of a floating dry dock causes three-dimension linear
displacements of any point of its wing deck. Extreme values of the According to our practice, if the length of the dock exceeds
linear deformations and typical requirements for deflection mea- 100 m, the minimum number of measurement points should be
suring systems are shown in Table 1. five or more, so three of them are placed as in the previous case (at
 V.V. Korotaev et al. / Ocean Engineering 117 (2016) 39–44
each end of the wind deck) and the rest two points are at 1/4 and
3/4 of the wing deck length. In this case, the fourth-degree poly-
nomial deformation profile can be calculated by solving a system
of five equations to find the polynomial coefficients k4 ; k3 ; k2 ; k1 ; k0 .
3. Camera-based deflection control system
3.1. Description of the system
The structure of the system to measure deflection of one wing
wall is shown in Fig. 3. The same set of units has to be mounted at
the second wing wall. The system to measure one wing wall
contains a set of reference marks, a measuring and a processing
units and a monitor. The reference marks are mounted at holders
at the measurement points along the wing deck to visualize the
linear displacement of these points. Electrical power is supplied to
each mark by wires. The measurement unit is placed in the middle
of the wing deck. The unit measures relative position of the marks
with respect to its own position. Each mark contains a set of
infrared light emitting diodes (LEDs) TSAL5100 (wavelength
940 nm, angle of half intensity 10°, optical power 35 mW) by
VISHAY. LEDs are projected to a camera sensor by a long focal-
length lens Jupiter 11 (f ¼ 135 mm, F set to 12). The camera is
based on Aptina MT9P031 sensor with 2592  1944 pixels. One
camera with lens forms “bow” measurement channel (Fig. 3),
another one forms “stern” channel. Each channel has an infrared
pass and visible cut off filter to match the spectral range of the
LEDs with the cameras. During the measurements the images from
bow and stern channels are analyzed by a designed software at the
processing unit, which is based on an industrial computer (945GSE
by LiPPERT). The information about current deflection is displayed
in a control room of the dock and transferred to a control system
of the dock. The overall system performance results in a mea-
surement data rate of 1 Hz.
In addition, the developed camera-based system (Fig. 3) does
not require significant infrastructure at a dock, so it can be
installed to a new or existing dock.
Fig. 3. Structure of designed and developed system for deflection measuring. Units to measure the deflection of one of two wing walls are depicted.
41
3.2. Principle of operation
A deflection of the dock causes linear shift of the reference
marks. As the marks are projected onto the cameras sensors, the
images with marks are captured and analysed to give the shift of
the marks. The developed software automatically detects the
marks images and measures their location at the cameras sensors.
The used algorithm for the detection of the mark image is pre-
sented in Fig. 4. First, an image with the switched on mark is
captured (Fig. 4(a)). A region of interest (ROI) is used to minimise a
search area (Fig. 4(b)). The ROI is defined during the installation of
the system and it corresponds to the defined measurement range
(see Table 1). Then an image with the switched of mark is captured
(Fig. 4(c)). As a result a frame-to-frame difference (Fig. 4(d)) which
reveals the mark location is calculated. In addition a check using
expected intensity level and the image size of the mark is done for
reliability.
Once the image of the mark is detected, the location of the
mark is measured in the raw image (Fig. 4(a)) with a subpixel
accuracy. For this task different kinds of algorithms with accura-
cies up to 0.01 pixel can be used (Zhukov et al., 2009; Mas et al.,
2012; Savransky et al., 2013; Haist et al., 2014). An iterative algo-
rithm based on Zhukov et al. (2009) was used due to its high
reliability. The detection and measuring algorithms are imple-
mented in an automatic mode, so no human interaction with the
software is required. The displacement hðzi Þ of the i mark located
at a distance zi can be found by estimating the coordinate of the
mark image ypix0i in pixels (Korotaev et al., 2010; Pantyushin and
Korotaev, 2012):
z
hðzi Þ ¼ ypix0i  p i0 ; ð3Þ
f
0 0
when zi ⪡f , where f is a back focal length of the lens, p is the
0
linear size of the sensor element (pixel). The ratio zi =f defines the
scale of the pixel to mm conversion. To simplify the estimation of
0
the distance zi and the back focal length f at a real dock the
calibration of the scale can be done with designed mark holders
(Fig. 3). The holders have mounting places for the marks at 0 and
7150 mm. The positions are predefined with accuracy up to
42 V.V. Korotaev et al. / Ocean Engineering 117 (2016) 39–44

deflection is

hmax ¼ ðh1 þ h2 ÞL2 =8z21 : ð7Þ

Considering a certain degree of correlation between errors of

measuring each mark location h1 and h2, these locations have to be
measured with error less than 7 2:5 mm to provide the required
7 5 mm accuracy of deflection hmax measurement (see Table 1).

4. Description of system tests

4.1. Laboratory test setup

Laboratory tests were done to verify and characterize the accu-

racy and measurement range of the developed system. According to
Fig. 4. Mark detection algorithm: (a) raw image with the mark and a background
safety and technical reasons it is not possible to verify the mea-
(including a dock illumination lamp), (b) cropped region of interest, (c) the back- surement range at a real floating dry dock, so a test setup with
ground when the mark is switched off, and (d) frame-to-frame difference. 100 m optical path was developed (Fig. 5).
In the middle of the path the measurement unit was installed.
70.1 mm. During the system installation the marks have to be Two reference marks were placed at each end of the path. Each
placed at these predefined positions and their coordinates at the mark was tightly mounted on a 300 mm holder, which was used
cameras sensors have to be measured. At the first stage of the for placing the marks in a vertical direction in predefined positions
installation process the marks are mounted at 0 positions. This with 50 mm step. Measurements of the marks position were done
allows determining initial deformation of the dock at unloaded
at each step within the measurement range. The results are pre-
state. The software remembers the measurement result and
sented in the next section.
assigns it to “zero” deformation. Then the marks are placed at
positions 7 150 mm. Obtained indications determine the scale for
each camera. The scale can be used for further measurements. 4.2. Field test setup
When displacements of all marks are measured, the overall
deformation profile of the wing wall can be calculated. Since the Field tests were done to verify the reliability of 24/7 operation
origin of the measurement coordinate system is at the measure- of the system in a real environment. The developed system was
ment unit, there is no offset of the deformation profile from this installed on a floating dry dock of 40,000 tons displacement in
point, so k0 (Eq. (1)) can be equaled to 0. In this case n  1 equa- Vladivostok, Russia. A total length of the dock was 125 m while
tions (Eq. (1)) are needed to define the deformation profile. total width of the dock wing walls was 39 m. Maximum allowable
According to the system structure (Fig. 3) the measurement operating deflection of the dock was 7100 mm.
unit observes opposite sides of the wing deck. Let us denote hn1 and During the tests one measurement unit and two reference
hm2 the linear shifts measured by bow and stern channels,
marks were used (Fig. 6) in order to measure the deflection. The
respectively (see Eq. (3)). In this case Eq. (1) can be rewritten as
reference marks were mounted at both ends of the wing wall of
2 4 3 2 13
z1 z31 z21 z1 h1 the dock. The measurement unit, the processing unit and the
6 72 3 6 7
6 ⋮ ⋮ ⋮ ⋮ 7 k4 6 ⋮ 7 monitor were placed at a control room of the dock, so full instal-
6 4 76 7 6 n 7
6 z 3 2
zn 7 k3 7 6 7 lation was done. Measurement results were presented to the dock
6 n zn zn 76 6 h1 7
6 4 76 7 7¼6 1 7 : ð4Þ
6 zn þ 1 z3n þ 1 z2n þ 1 zn þ 1 76 k 6 7 crew. The tests lasted for three weeks of continuous operation,
6 74 5 6 2 7
2 h
including occurred rain, fog and windy weather conditions.
6 ⋮ ⋮ ⋮ 7
⋮ 5 k1 6 7
4 4 ⋮ 5
As a final stage of the field tests a stability of the system's
z4n þ m z3n þ m z2n þ m zn þ m m
h2
indication was checked. First, we checked that the measurement of
In practice, an assumption that the deflection of a dock is the dock deformation still indicated “zero” for both channels when
symmetrical to its middle can be done, so the deflection profile can the dock was unloaded. After that the marks were placed to the
be characterized by a parabola with origin in the dock middle mounting positions 7 150 mm. This proved that the scale of the
(Hedger, 2007). In this case two reference marks are needed. If the channels did not change too during the field tests.
distances between them and the measurement unit are the same
ðj z1 j ¼ j z2 j Þ, Eq. (4) can be simplified to two equations:
2
h1 h2
6 k2  z2 þ k1  z1 ¼ h1 k1 ¼
6 1 2z1
6 ) h1 þh2 ð5Þ
4 k2  z21  k1  z1 ¼ h2 k2 ¼ 2
2z1

Hence, full deflection profile hðzÞ (see Fig. 2) can be described as

h1 þ h2 2 h1  h2
hðzÞ ¼ z þ z: ð6Þ
2z21 2z1

Maximum magnitude of the dock deflection hmax with length L

can be characterized at the endpoints of the wing deck at bow
zend1 ¼ L=2 and stern zend2 ¼  L=2, so the magnitude of the Fig. 5. A laboratory setup designed for the system tests.
 V.V. Korotaev et al. / Ocean Engineering 117 (2016) 39–44 43

Fig. 6. Developed deflection measurement system installed on a 125 m floating dry dock during the field tests.

Fig. 7. Results of the laboratory tests: (a) and (b) error of measurement of the bow and stern marks position, respectively, and (c) the resultant error of the deflection
measurement.

resultant error of the deflection measurement (Fig. 7(c)) is less

Table 2
Technical specifications of the designed camera-based deflection measuring
system.
Parameter
Measuring range of vertical shift of the reference mark,
mounted 507 5 m away from the measurement unit
Resolution
Accuracy of vertical shift measuring of the reference mark,
mounted 507 5 m away from the measurement unit
Warm-up time
5. Results and discussion
Experimental results obtained during laboratory tests are
shown in Fig. 7. As one can see, measuring error of the bow and
stern marks positions (Fig. 7(a) and (b), respectively) is less than
71:5 mm within the measurement range of 7 150 mm. The
than 71.5 mm, that is three times smaller than required 75 mm
(see Table 1). Proved parameters of the developed system are
presented in Table 2.
Value During laboratory tests we also tested the measurement unit
for additional shock and vibration loads at 0.5–3.5 Hz with 0.5 g
o 7 150 mm
magnitude, that proved its ability to keep the optical alignment at
70.1 mm harsh environment.
71.5 mm Field tests at real floating dry dock infrastructure proved that
the system is capable to work within temperature range from
o 5 min
 15 °C to þ 45 °C and relative humidity up to 100%. All units of
the system have protection level against water, dust and sand no
less than IP 66. During field tests all parts of the system also
proved resistance to rain (up to 1 mm/h), salt fog and airflow with
velocity up to 25 m/s. There were no breakdowns during all tests
and operation period.
In addition the developed system passed the assessment of
Russian Maritime Register of Shipping (Rules for Technical
Supervision during Construction of Ships and Manufacture of
44 V.V. Korotaev et al. / Ocean Engineering 117 (2016) 39–44
Materials and Products for Ships, Part V Technical Supervision
During Construction of Ships, 2014) on international requirements
for deflection measuring systems (Germanischer Lloyd Aktienge-
sellschaft. Rules for Classification and Construction. Floating
Docks, 1993; Rules for Classification of Floating Docks, 2009; Rules
for Building and Classing. Steel Floating Dry Docks, 2009).
It should be noted, that the system is capable of measuring
linear shift of reference mark along X 0 direction too (see Fig. 1).
This is secondary data in general dock operation, but it can be
presented without changes of system structure (owning to used
camera sensors).
For future development additional improvements of the system
can be done. A deflection profile of a floating dry dock depends on
a water–air temperature difference (Tupper, 2013). So additional
temperature compensation of the measurement output can be
implemented. Another challenge for the deflection measurement
by optics is the sun heating of the wing deck, that causes air
refraction and turbulence. Still it is important to notice that
combined influence of the said factors does not exceed 1 mm at
working distances and the total error of the system taking these
factors into account is still smaller than required 75 mm.
6. Conclusions
In this paper a system for deflection measuring of floating dry
docks is presented. The developed system is the camera-based
device that employs digital image processing of the spatial posi-
tion of reference marks placed along each wing deck of the dock.
Approach to perform the deflection control of the floating dry
dock based on measuring of position of a finite number of points
along the wing deck was described. The system can be installed to
a new or an existing dock and does not require significant infra-
structure. Test results of use of the developed system both in the
laboratory facility and the real dock infrastructure are provided
and discussed. The laboratory tests indicated that the developed
system demonstrates good measurement accuracy and the tests at
the real dock showed that the system is capable to work in a harsh
environment with no breakdowns.
Acknowledgements
This work was partially financially supported by the Govern-
ment of the Russian Federation, Grant 074-U01.
References
Carbonari, S., Gara, F., Roia, D., Leoni, G., Dezi, L., 2013. Tests on two 18-years-old
prestressed thin walled roof elements. Eng. Struct. 49, 936–946.
Froggatt, M., Moore, J., 1998. High-spatial-resolution distributed strain measure-
ment in optical fiber with rayleigh scatter. Appl. Opt. 37 (10), 1735–1740.
Gaythwaite, J.W., 2004. Design of Marine Facilities for the Berthing, Mooring, and
Repair of Vessels, 2nd edition.
Germanischer Lloyd Aktiengesellschaft. Rules for Classification and Construction.
Floating Docks, 1993.
Gorbachev, A.A., Konyakhin, I.A., Musyakov, V.L., Timofeev, A.N., 2007. Study of the
structural features of invariant optoelectronic systems with a unified matrix
analysis field. J. Opt. Technol. 74 (12), 810–814.
Haist, T., Dong, S., Arnold, T., Gronle, M., Osten, W., 2014. Multi-image position
detection. Opt. Express 22 (12), 14450–14463.
Hedger, R., 2007. Dockmaster Training Manual.
Kleshchenok, M.A., Anisimov, A.G., Lashmanov, O.U., Timofeev, A.N., Korotaev, V.V.,
2014. Alignment control optical-electronic system with duplex retroreflectors.
In: SPIE Photonics Europe. International Society for Optics and Photonics,
pp. 91311X–91311X.
Korotaev, V.V., Konyakhin, I.A., Timofeev, A.N., Yaryshev, S.N., 2010. High precision
multimatrix optic-electronic modules for distributed measuring systems. In:
Sixth International Symposium on Precision Engineering Measurements and
Instrumentation. International Society for Optics and Photonics, pp. 75441E–
75441E.
Korotaev, V.V., Timofeev, A.N., Gorbachev, A.A., Pantjushin, A.V., Aleev, A.M., Kule-
shova, E.N., 2012. Apparatus for controlling deformations of elongated object.
RU Patent 2445572, March 20.
Korotaev, V.V., Timofeev, A.N., Ivanov, A.G., 2000. Problems in the development of
optoelectronic systems for monitoring displacements of large-sized objects. J.
Opt. Technol. c/c Opticheskii zhurnal 67 (4), 336–339.
Lynch, J.P., Wang, Y., Loh, K.J., Yi, J.-H., Yun, C.-B., 2006. Performance monitoring of
the geumdang bridge using a dense network of high-resolution wireless sen-
sors. Smart Mater. Struct. 15 (6), 1561.
Maraev, A.A., Timofeev, A.N., 2013. Energetic sensitivity of optical-electronic sys-
tems based on polychromatic optical equisignal zone. In: SPIE Optical Metrol-
ogy 2013. International Society for Optics and Photonics, pp. 878836–878836.
Maraev, A.A., Vasilev, A.S., Timofeev, A.N., 2014. Study of irradiance distribution in
optical equisignal zone. In: SPIE Photonics Europe. International Society for
Optics and Photonics, pp. 91380Q–91380Q.
Mas, D., Ferrer, B., Sheridan, J.T., Espinosa, J., 2012. Resolution limits to object
tracking with subpixel accuracy. Opt. Lett. 37 (23), 4877–4879.
Mott, R.L., 2014. Applied Strength of Materials. CRC Press.
Newman, T.S., Jain, A.K., 1995. A survey of automated visual inspection. Comput.
Vis. Image Underst. 61 (2), 231–262.
Nixon, M., Nixon, M.S., Aguado, A.S., 2012. Feature Extraction and Image Processing
for Computer Vision. Academic Press.
Pantyushin, A.V., Korotaev, V.V., 2012. Control measurement system for railway
track position. In: SPIE Optical Engineering þ Applications. International
Society for Optics and Photonics, pp. 84861B–84861B.
Rules for Building and Classing. Steel Floating Dry Docks, 2009. American Bureau of
Shipping.
Rules for Classification of Floating Docks, 2009. China Classification Society.
Rules for Technical Supervision during Construction of Ships and Manufacture of
Materials and Products for Ships, Part V Technical Supervision During Con-
struction of Ships, 2014. Russian Maritime Register of Shipping.
Sannerhaugen, V., Hellvik, S., 1999. Method and device for continuous monitoring
of dynamic loads. US Patent 5,942,750, August 24. URL: 〈http://www.google.
com/patents/US5942750〉.
Savransky, D., Thomas, S.J., Poyneer, L.A., Macintosh, B.A., 2013. Computer vision
applications for coronagraphic optical alignment and image processing. Appl.
Opt. 52 (14), 3394–3403.
Smith, D., LeVezu, A., 2012. Floating dock deflection management systems. US
Patent 8,155,812, April 10. URL: 〈https://www.google.com/patents/US8155812〉.
Smith, W.F., Hashemi, J., 2006. Foundations of Materials Science and Engineering.
Mcgraw-Hill Publishing.
Sonka, M., Hlavac, V., Boyle, R., 2014. Image Processing, Analysis, and Machine
Vision. Cengage Learning.
Stiros, S.C., Psimoulis, P.A., 2012. Response of a historical short-span railway bridge
to passing trains: 3-d deections and dominant frequencies derivedfrom robotic
total station (rts) measurements. Eng. Struct. 45, 362–371.
Swartz, R.A., Zimmerman, A.T., Lynch, J.P., Rosario, J., Brady, T., Salvino, L., Law, K.H.,
2012. Hybrid wireless hull monitoring system for naval combat vessels. Struct.
Infrastruct. Eng. 8 (7), 621–638.
Tupper, E.C., 2013. Introduction to Naval Architecture: Formerly Muckle's Naval
Architecture for Marine Engineers. Butterworth-Heinemann.
Vincent, P.S., Gardiner, C.P., Wilson A.R., Ellery, D., Armstrong, T., 2008. Installation
of a sensor network on an ran armidale class patrol boat. In: Materials Forum,
vol. 33, pp. 307–316.
Yang, G., Liang, H., Wu, C., 2013. Deflection and inclination measuring system for
floating dock based on wireless networks. Ocean Eng. 69, 1–8.
Zhukov, D.V., Konyakhin, I.A., Usik, A.A., 2009. Iterative algorithm for determining
the coordinates of the images of point radiators. J. Opt. Technol. 76 (1), 36–38,
URL: 〈http://jot.osa.org/abstract.cfm?URI ¼jot-76-1-36〉.
Zou, L., Bao, X., Yang, S., Chen, L., Ravet, F., 2006. Effect of brillouin slow light on
distributed brillouin fiber sensors. Opt. Lett. 31 (18), 2698–2700.