International Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 5. Hakodate 1998
THE EXTRACTION OF 2-D SEA PARAMETER AT NEAR SHORE REGION FROM SEQUENCE
ANALYSIS OF MARINE RADAR IMAGE AND ITS VERIFICATION S(k
Shintaro GOTO, and Yoshibumi MIZUNO** Th
i : ) e
Associate Professor, Environmental Information Laboratory Wal
Kanazawa Institute of Technology We
7-1 Ohgigaoka Nonoichi Kanazawa Ishikawa 921, JAPAN ast
E-mail: got env.kanazawa-it.ac.jp
** Mitsui Zosen Systems Research Inc. «
SCH
Commission V, Working Group IC WG V/III
Wh
Wa
KEYWORDS: Image sequence analysis, Optical flow model, Marine radar, Wave parameter wav
sms
wav
ABSTRACT
NEN Sue | | L=
The Observation of 2-D wave parameter distribution is important for estimating the wave force to marine constructions,
such as jetty. The in situ data routinely observed now are point data and the cost is expensive, so the number of whe
observation point is a few. There is no routine method for monitoring 2-D wave parameter distribution.Marine radar can dep
offer the image of 2-D wave field by monitoring the sea clutter from the sea surface.
In this study we observed the wave parameter around the coast area near Komatsu city besides Sea of Japan by marine
radar from Jan.8 to Jan.11, 1996 when the highest significant wave in winter of 1996's was observed. From Image of
marine radar we obtained the wave direction and wave speed by optical flow estimation. Wave direction and wave speed
were calibrated by the result of FFT and in situ data by Ministry of Construction Japan. The Accuracy of wave direction
fits to the in situ data. But the accuracy of wave frequency is lower than wave direction. To examine this correlation, the
nonlinearity caused by wave breaking was investigated with the simulation study using in situ measurements.
1. INTRODUCTION
The Observation of 2-D wave parameter distribution is
important for the estimation of the wave force to marine
constructions, such as jetty. The in situ data routinely
observed now are point data and the cost is expensive, so
the number of observation point is a few. There is no
routine method for monitoring 2-D wave parameter
distribution.
Marine radar can offer the image of 2-D wave field by
monitoring the sea clutter from the sea surface.
F.Ziemer,and W.Rosenthal!) was developed the transfer
function for marine radar for imaging ocean waves. After
applying the transfer function to the image spectra,
significant wave heights were estimated and compared
with results from other sensors during LEWEX2). Results
of their study are very promising. And M.D.Henschel et
al3) developed the prototype marine radar ocean wave
imaging system, MACRADAR, for validation of the ERS-1
SAR wave mode operation by using F. Ziemer et al 's
method. But their application was carried out only on the
offshore region.
In this study we observed the wave parameter around the
coast area near Komatsu city besides Sea of Japan by
marine radar from Jan.8 to Jan.11, 1996 when the highest
significant wave in winter of 1996's was observed. From
Image of marine radar we obtained the wave direction
and wave speed by optical flow estimation. Wave
direction and wave speed were calibrated by the result of
FFT and in situ data by Ministry of Construction Japan.
698
The Accuracy of wave direction fits to the in situ data. But
the accuracy of wave frequency is lower than wave
direction.
2. METHODOLOGY OF WAVE PARAMETER
DETECTION
2.1 Wavelength and Wave Direction by FFT
Fig.1 shows the flow diagram of analysis procedure. As
the received power from the sea surface become more
and more weakly as the radar is far from the radar site.
The correction function are applied to correct the
difference of these power due to the distance from the
radar to the sea surface. And the received power from the
radar is PPI image shown as Fig.2 on polar coordinate
system, so the image is transformed to x-y coordinate
system for FFT analysis. The Fourier transformation
F(kx,ky) is defined by
oc
F(kxky)=1/(27)2 + [7 (xy) *
Ex
exp(-i(kxx+kyy))dxdy |. su. 1)
2(Xy) is the radar image defined on x-y coordinate
System. And Wave number spectrum is defined by