Transcript 3.1. Video-derived buoy positions

Management of dynamic navigation
channels using video techniques
(the Teignmouth case)
Mark Davidson – University of Plymouth, UK
Ismael Mariño-Tapia - University of Plymouth, UK
The CoastView Project
Contents
1. Site description
• General (Location, economic activities, wave climate).
• Morphodynamics
2. Problem description
3. Video solution
• Frame of reference & CSIs
3.1. Video-derived buoy positions
• Algorithm (Methods)
• Validation of the technique
• Results
3.2. Video-derived dredging tracks and dredging
intensity
• Algorithm
•Results
4. Conclusion
Site Description..(1)
• General
• Located in the coast of South Devon, UK.
• Fronted by 2km beach, to the south meets the Teign.
• Infrequent wind-generated wave climate from NE to S
• Protected from Atlantic swell
• Macrotidal conditions (STR ~4.2m)
• Economic activity
• Dock, port and harbour 13% of total
economic activity
Site Description..(3)
Morphodynamics (updated version of cycle)
Stage II
Stage I
Beach
Shore-attached
sand bar
Ac
ss
ce
Ch
n
an
Offshore sandbank
(North)
el
A
es
cc
s
C
n
ha
ne
l
Offshore sandbank
(South) / Ness pole
Ness
Head
Stage III
Key
Beach
•
ds
an
tS
rat
Sp
Offshore sandbank
(North)
ss
ce
c
A
C
n
ha
l
ne
Offshore sandbank
(South) / Ness pole
Limit of intertidal region
Subtidal contour (-3 m ODN)
Sediment transport pathways
Dry land
Scale:
200 m
Problem Description
•
Due to dynamic nature of sanbanks the channel entrance
is allowed to move within limits.
• Dredging (ploughing) occurs at will to maintain channel
entrance, it has proven difficult to maintain an effective
dredging strategy.
•Difficult to position buoys adequately relative to channel
& hazardous sandbanks
Video solution
“Improve navigation safety and avoid human,
economical or ecological damage by minimizing
the possibility of ships going aground”
“To ensure that the buoys accurately mark the
channel perimeter”
1.CSI: Sandbar-Buoy
Inter-distance (SBID)
2. Benchmarking
exceeded
SBID should not exceed X
m / Invade buffer region
Current SBID
“Preserve integrity of channel entrance by
effective dredging operations”
3. Move buoys
4. Evaluation procedure
Video solution
“Improve navigation safety and avoid human,
economical or ecological damage by minimizing
the possibility of ships going aground”
Morphological evolution
(including subtidal)
“Preserve integrity of channel entrance by
effective dredging operations”
Contribution of natural
processes (model?)
Video-derived variable:
Dredging intensity
Benchmarking
exceeded
“Innapropriate” channel
Assess effects of dredging
(CSI)
Current channel entrance
location/depth
Change dredging
strategy
Video-derived buoy positions
Methodology (Algorithm)
1.
Reduce the search area
2.
Isolate red band
3.
Detection of buoy
C  max ii 1n f i  min( f i )
[u, v]  find ( f  C )
4.
Transform oblique coordinates to
planview (UV to XYZtide)
Video-derived buoy positions
Validation of the technique
•
Buoy position measured in June 03 and 04 with total station (24 hrs).
•
Errors: X-shore = 1 m (0.25 resolution); L-shore = 2 m (1 m resolution)
Cross-shore position (Argusx)
Along-shore position (Argusy)
264
262
226
260
224
258
256
07:00
13:00
19:00
Measured
Video-derived
222
220
07:00
13:00
19:00
Video derived buoy positions
Results (images)
• Images: Permit the identification of buoy position relative to
dangerous sandbanks and present the displacement of buoy
positions in 2D.
Video derived buoy positions
Results (time series)
• Time series permit the identification of subtle trends related to
natural drift of the buoy, and to assess the effects of individual events
(storms and tides) on variability of the displacement.
370
360
350
340
330
320
Time series of cross-shore position for buoy 2
380
Cross-shore position from estuary (m)
Cross-shore position from estuary (m)
380
375
370
365
360
355
350
345
340
310
Buoy X-shore movement
Tides
Waves
Sep
Oct
Nov
Jan Feb MarchApril May June July Aug Sep Oct Nov Dec
Dec
Video derived buoy positions
Defining benchmarks
• Sandbank-buoy interdistance (SBID) works as a safe CSI only if
morphology below the reference isobath does not change
substantially or has a constant slope.
Video derived buoy positions
Defining benchmarks
0
Channel profiles at X=180
-1
-2
-3
May-99
Oct-99
Nov-99
Aug-02
Nov-02
May-03
Nov-03
Jun-04
-4
-5
-6
50
100
150
200
250
300
0
Channel profiles at X=40
-1
-2
-3
-4
-5
-6
-7
10
60
110
160
210
350
400
Video-derived dredging tracks and
intensity
Algorithm (Mathematical morphology)
Dredging
track
1.
Convert to grey
Identify
objects in
2. 1.Correct
uneven
common (sandbanks)
illumination
(tophat
filtering)
2. Filter them out using
image
segmentation
3. Apply
threshold
Filter other noise
4. 3.Label
sources not related Submerged
to
interconnected
sandbanks
elements
as (size, Sandbank
distanceobjects
form centre of
individual
mass, etc).
Video-derived dredging tracks
and
intensity
Results
•Location of the dredging tracks, the intertidal morphology and the
presence of submerged sandbanks (breaking patterns).
•Regions where dredging effort is focused can be identified
Environmental parameters from Mon 3 May to Fri 21 May
2
0
-2
1.5
1
Tidal curve
Video-derived dredging tracks
and
intensity
Wave height
Results (example)
0.5
0
03/05
05/05
07/05
09/05
11/05
13/05
15/05
17/05
19/05
21/05
Concluding remarks: Navigation

Video-derived buoy positions




Determine the position of navigation buoys relative to
hazardous sandbanks.
Track their movement towards a dangerous position.
Accuracy within 1 m in the cross-shore and 2 m in the
longshore.
Benchmarking : SBID from -2 m ODN with safety
margins in case of ‘step’ on sub tidal morphology.
Concluding remarks: Navigation

Video derived dredging intensity



Able to determine the position of dredging and the
intensity (areas of focus) in relation to sandbanks.
Not able to assess the effects of dredging on its own.
Very useful video derived variable if used in
conjunction with extra information.
Thanks for your attention