Transcript Mushroom, fish and Classification machines With focus on linear

Monitoring fish
migration in Rivers
Helge Balk
Department of Physics. University of Oslo.
1
River Tana
2
River Tana
3
Tornio (Se/Fi)
4
River Reisa
(north Norw.)
5
River Fischa in
Austria
6
Echosounder in the river
Amplitude
Detector
4-Ch
TVG
1 size
correction
Phase
detector
7
What we can do to increase the
detection probability
Estimate the DPF
Reduce the variation in the DPF
Site selection
Bottom modification
Surface modification
Guide fish
Opening angle
Sound frequency
8
If we detect a fish, What does it
mean?
Echo sounder
Fish detection software
?
9
Interpretation: DPF, noise and time
expansion
Echo
sounder
Interpreter
Fish
detector
Meaningful
statistics
Measure noise level as a function of time and range over years
Apply noise level as an overall fish size threshold
Use noise level to correct the number of fish within each size group
Apply time expansion to find the total fish abundance estimate
Intensity
Halt in
recording
High noise
level
High noise
level
Noise level
Threshold
1
2
3
4
5
6
7
8
9
10
Time
10
0
Interpretation:
Area expansion
Echo
sounder
1
Covered area
Uncovered area
Fish
detector
1
2
2
3
!
Interpreter
Meaningful
statistics!
What did we cover?
Where did the fish pass?
Index or total run?
11
Can we estimate TS in a river?

Spherical or cylindrical spreading?
Corrupted vertical angular measurements?
Additive noise?
Can we compensate for swimming motion?

Can we compensate for fish aspect?



12
Can we estimate TS in a river?
Alternative methods to establish the size?
Beam intensity mapping?
Apply reference targets?
Multiple narrow beams?
Multiple frequencies?
Aspect detection and correction?
13
x
14