Television – Camera – system and detail – Combined image signal generator and general image signal...
Reexamination Certificate
1996-07-31
2003-02-04
Moe, Aung S. (Department: 2612)
Television
Camera, system and detail
Combined image signal generator and general image signal...
C348S674000, C348S222100
Reexamination Certificate
active
06515699
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to video camera processing and, more particularly, to preventing aliasing in video camera signals.
A known digital video camera is shown in FIG.
29
and generates video camera signals which are subject to aliasing. Aliasing is that phenomenon which occurs when an analog signal is digitally sampled at an insufficient sampling rate. The lowest sampling rate which produces a sampled signal that can be reconverted to the original analog signal is known as the Nyquist frequency or rate. The resulting aliased signal is a lower frequency version, or “alias”, of the original signal. Aliasing is particularly a problem when a non-linear function, such as a gamma correction function (FIG.
2
A), is applied to the video signal because the correction function adds high frequency harmonics to the input video signal. These high frequency harmonics require a higher sampling rate than anticipated and, therefore, produce alias signals.
The digital video camera of
FIG. 29
is affected by aliasing because it includes a gamma correction circuit
134
as part of its video camera signal processing. An optical system
100
provides a focussed image which is converted into a video signal by a charge coupled pickup or sensing device
110
(CCD). The video signal is, then, pre-amplified by a pre-amplifier
111
and video amplified by a video amplifier
112
. The video amplified signal is digitized by an analog-to-digital converter
113
and forwarded to a defect correction circuit
114
for digital correction. The corrected video signal is delayed by a first delay circuit
115
and, then, further delayed by a second delay circuit
116
. The twice delayed video signal is applied to a linear matrix
132
for correcting a color reproduction error, which arises because the photographing performance of the CCD in reality is different from an ideal photographing performance. After being combined with an image contour signal by an adder
130
, the linearized video signal is applied to a series of correction circuits, which includes a knee correction circuit
133
, the problematic gamma correction circuit
134
and a B/W clip circuit
135
. The gamma correction circuit
134
applies a non-linear function to the video signal which gives rise to the aliasing problem.
The gamma correction circuit of
134
receives digital samples of the video signal at a sampling rate f
S
and converts each received digital sample into a value which best fits the gamma correction function shown in FIG.
30
. That is, the gamma corrected signal is not ideal and results in a collection of values which are approximations of the ideal. When this occurs, unwanted frequency harmonics are produced by the gamma correction function. Where the frequency of the input video signal is high, the sampling rate f
S
may not be sufficient to accurately represent the input signal. Thus, sampling occurs at a lower rate than the Nyquist rate and aliasing occurs.
The aliasing problem is graphically illustrated by
FIGS. 31A-31D
which show the harmonics in the frequency domain. Aliasing occurs when the harmonics overlap with complement harmonics as shown in FIG.
31
B. An ideal sinusoidal wave has a single harmonic f and, therefore, yields a sinusoidal wave at the output of an ideal gamma correction circuit represented by the frequency component at frequency f shown in FIG.
31
A. However, the usual gamma correction circuit is not ideal and produces the harmonics shown in
FIG. 31B
which are produced at frequencies according to the asymptotic function of sampling theorem. The original signal can be reconstructed only so long as the frequency f is low and the harmonics do not substantially overlap with the harmonics of the complement signal at f′, as shown in FIG.
31
C. However, aliasing occurs when the frequency f of the video signal is high and shifts closer to its complement frequency f′. In this situation, as shown in
FIG. 31D
, the harmonics overlap and are combined and, therefore, the resulting digital signal yields an aliased analog signal which cannot be reconstructed into the original video signal (FIG.
31
D).
Harmonics also arise when image contour processing is applied to the video signal. For example, in the video camera of
FIG. 29
, an image contour is emphasized by processing the video signal in the horizontal and vertical directions after the video signal has been delayed by delay circuits
115
,
116
and
117
. A vertical direction high pass filter (HPF)
121
and a horizontal direction low pass filter (LPF)
122
function to pass the vertical direction component of the image contour signal to a multiplier
123
. Similarly, a vertical direction LPF
124
and a horizontal direction HPF
125
pass the horizontal direction component of the image contour signal to a multiplier
127
. The vertical and horizontal contour signals are multiplied by respective gain adjustment signals applied to respective terminals
144
and
145
to emphasize the contours in the multipliers
123
,
127
, respectively. The emphasized vertical and horizontal contour signals are combined by an adder
128
to form the emphasized image contour signal which is fed to a limiter
129
for limiting the output of the adder
128
such that the resultant limited signal is not overly emphasized.
The image contour processing also generates high frequency components which give rise to the aliasing problem. More specifically, when the gamma correction circuit
134
digitally samples the contour signals contained in the output of adder
130
and which contains high frequency components, aliasing occurs and the original contour signal cannot be reconstructed.
Although the problem of aliasing which arises from contour image processing would be avoided if the contour image signal is combined with the video signal after gamma correction, another problem arises because the gamma correction function serves to amplify the video signal. Therefore, if the contour image signal is combined with the video signal after gamma correction, the contour image signal is relatively small as compared with the amplified video signal. As a result, the contour of an image is not adequately represented in the displayed video picture. Thus, it is not a sufficient solution to combine the image contour signal with the gamma corrected video signal after gamma correction.
The problem of aliasing will be further explained with reference to
FIG. 32
which schematically depicts a simplified configuration of the video camera shown in
FIG. 29
, and in which a video signal is received at input terminal
160
and digitized by an analog-to-digital converter
161
to produce the digitized video signal (a
S
) of FIG.
33
. The digitized video signal (a
S
) is output to a high pass filter
162
(corresponding to the contour image processing circuitry) and to a low pass filter
164
(corresponding to the linear matrix
132
). The image contour processed signal (b
S
) of
FIG. 34
is combined with the linearized video signal (c
S
) of
FIG. 35
by an adder
168
to yield the video signal with emphasized contours (d
S
) of FIG.
36
. The emphasized video signal (d
S
) is fed to a gamma correction circuit
167
which produces the gamma corrected signal (e
S
) of
FIG. 37
at an output terminal
169
.
It will be noted from
FIG. 33
that the signal (a
S
) includes several frequency harmonics which are filtered by the high pass filter
162
, resulting in the image contour signal (b
S
) shown in
FIG. 34
having the low frequency components removed. Conversely, the low pass filter
164
, representing the linear matrix
132
(FIG.
29
), filters out high frequency components and results in the linearized signal (c
S
) of
FIG. 35
having its high frequency components removed. The combined signal (d
S
) shown in
FIG. 36
is the sum of the image contour signal (b
S
) and the linearized signal (c
S
). At this point, it will be noticed that the combined signal (d
S
) includes several frequency components which is indicative of the frequency modulate
Kihara Taku
Sudo Fumihiko
Tanji Ichiro
Frommer William S.
Frommer & Lawrence & Haug LLP
Moe Aung S.
Sony Corporation
LandOfFree
Anti-aliasing video camera processing apparatus and method does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Anti-aliasing video camera processing apparatus and method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Anti-aliasing video camera processing apparatus and method will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3160260