Transcript colin_B_squid

Image Processing for HTS
SQUID probe microscope
Advanced Image Processing
Seminar
Colin Bothwell
0570063
General Discussion of Paper




Use High Temp SQUID probe microscope
Probe allows high spatial resolution of
measurement of samples even at room
temp
Aim to produce magnetic field images
Improve images using Image Processing
Theory of SQUIDs



Superconducting QUantum
Interference Device
Measure tiny magnetic fields (threshold
is around 10-14 T where
1T=10000Gauss. )
Magnetic field of heart=10-10T
Theory of SQUIDs



To understand
SQUIDs need to
know basic
principles
Superconductivityno resistance
Josephson Effectcurrent flow carried
by cooper pairs
Theory of SQUIDs


Josephson Junctiona junction of 2 S/C
materials divided by
non-S/C material
(oxide)
With these
principles, SQUID
was designed
Theory of SQUIDs




SQUID has two
Josephson Junctions
Apply current to
SQUID
Voltage will oscillate
Oscillations depend
on magnetic flux
Experimental Set-Up


Developed HTS
SQUID probe
microscope using
fine permalloy
needle as flux guide
Images the
magnetic field by
means of raster scan
Experimental Set-up


Separation of probe
and SQUID adjusted
by viewing through
glass window
PC used to control
SQUID
Magnetic Image-Deficiencies





Resultant images are noisy and unclear
because of
Background Noise
Drifting
Data Jumping
So…
Image Processing!





Background Noise
Use image processing
algorithm
Calculate mean and
standard deviation of
power spectrum
Threshold value=Pm +
αPs
Apply this to each scan
line in same way
Drift between Lines



Drift of flux bias
point of SQUID
Treated with another
algorithm
Drift calculated as
difference in the
mean value of
magnetic field for
each scan line
Data Jumping


Artifacts due to a
jump in the flux bias
point in SQUID by
flux trapping or
unexpected noise
Jump causes change
dynamic range of
magnetic field to
become larger
Data Jumping


Small changes can no longer be detected
IP detects line with jump by derivation of
line data (higher derivation considered to
contain jump) so is interpolated with
neighbouring lines
Conclusion



Used high-permeability needle to enable high
spatial resolution measurement of samples.
Developed IP algorithms for magnetic field
data obtained from SQUID probe microscope.
Algorithms can remove problems of
background noise, drifting and jump data.