09-07-2012, 02:27 PM
Line Pattern Removal and Enhancement Technique for Multichannel
Passive Millimeter Wave Sensor Images
[attachment=27272]
INTRODUCTION
The detection, acquisition, classification, aim
point selection of ground mobile, high value targets
in high clutter, and adverse weather environment
are critical issues in development of smart weapons.
Advanced signal/image processing algorithms will
be the major element for the success of the smart
weapons. A significant problem that affects the
successful realisation of the goals of many tactical
missions is the poor resolution of images collected
from the sensors used to assist guidance operations.
INTERCHANNEL NOISE REMOVAL TECHNIQUES
A major problem of single-channel PMMW
imaging systems is the slow response time due to
the lack of thermal sensitivity. The imager could
be operated at TV (television) rates to reduce the
extent of this problem using a number of parallel
channels. Any PMMW imager that is able to operate
at a frame rate comparable with a thermal imager
will require several detectors that are scanned or
a starting array.
Noise-Removal Technique
Bernstein,12 et al. developed the noise-removal
technique to remove both kinds of noises–striping
and a herring bone noise pattern, due to coherent
noise from the satellite's 32 kHz switching power
supply superimposed on the detector signal. There
are the following two methods to remove a coherent
noise source of a known frequency:
RESULTS AND DISCUSSION
To test the efficiency of these techniques,
several data are considered with different sizes
(from 50 x 50 pixels to 450 x 450 pixels). Figure
2(a) shows an original image acquired by 1 x 16
multichannel PMMW sensor at 94 GHz and one
can see the visual distortions in the image due to
difference in inter-channel responses. It is necessary
to correct this inter-channel noise before further
enhancing the images. Figure 2(b) shows the result
by applying the frequency domain line pattern removal
technique, which is based on DFT technique in the
Fourier domain.
Passive Millimeter Wave Sensor Images
[attachment=27272]
INTRODUCTION
The detection, acquisition, classification, aim
point selection of ground mobile, high value targets
in high clutter, and adverse weather environment
are critical issues in development of smart weapons.
Advanced signal/image processing algorithms will
be the major element for the success of the smart
weapons. A significant problem that affects the
successful realisation of the goals of many tactical
missions is the poor resolution of images collected
from the sensors used to assist guidance operations.
INTERCHANNEL NOISE REMOVAL TECHNIQUES
A major problem of single-channel PMMW
imaging systems is the slow response time due to
the lack of thermal sensitivity. The imager could
be operated at TV (television) rates to reduce the
extent of this problem using a number of parallel
channels. Any PMMW imager that is able to operate
at a frame rate comparable with a thermal imager
will require several detectors that are scanned or
a starting array.
Noise-Removal Technique
Bernstein,12 et al. developed the noise-removal
technique to remove both kinds of noises–striping
and a herring bone noise pattern, due to coherent
noise from the satellite's 32 kHz switching power
supply superimposed on the detector signal. There
are the following two methods to remove a coherent
noise source of a known frequency:
RESULTS AND DISCUSSION
To test the efficiency of these techniques,
several data are considered with different sizes
(from 50 x 50 pixels to 450 x 450 pixels). Figure
2(a) shows an original image acquired by 1 x 16
multichannel PMMW sensor at 94 GHz and one
can see the visual distortions in the image due to
difference in inter-channel responses. It is necessary
to correct this inter-channel noise before further
enhancing the images. Figure 2(b) shows the result
by applying the frequency domain line pattern removal
technique, which is based on DFT technique in the
Fourier domain.