Television – Format conversion – Changing number of lines for standard conversion
Reexamination Certificate
1999-09-13
2002-08-13
Kostak, Victor R. (Department: 2711)
Television
Format conversion
Changing number of lines for standard conversion
C348S441000
Reexamination Certificate
active
06433828
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a picture information converting apparatus and a picture information converting method.
2. Description of the Related Art
In high audio-visual environments, television receivers that can display pictures with high resolution have been desired. To satisfy that, a so-called high-vision (high definition television) system has been developed. In the high-vision system, the number of scanning lines is 1125 that is more than twice of that of the conventional NTSC system. In the high-vision system, the aspect ratio—the ratio of the frame width to the frame height is 9:16 that is wider than 3:4 of the NTSC system. Thus, in the high-vision system, pictures that have high resolution and presence can be obtained.
When an NTSC format picture signal is supplied to a high-vision receiver, it cannot display a picture due to the difference of the signal format. To solve such a problem, with a picture information converting apparatus as shown in
FIG. 1
, the rate of the picture signal is converted. Referring to
FIG. 1
, an NTSC format picture signal as an SD (Standard Definition) signal is input from an input terminal
100
to a horizontal interpolation filter
101
. The horizontal interpolation filter
101
performs a horizontal interpolation process for the NTSC format picture signal. The output signal of the horizontal interpolation filter
101
is supplied to a vertical interpolation filter
102
. The vertical interpolation filter
102
performs a vertical interpolation process for the output signal of the horizontal interpolation filter
101
. The vertical interpolation filter
102
outputs a high-vision format picture signal as an HD (High Definition) signal.
Next, with reference to
FIG. 2
, a practical structure of the horizontal interpolation filter
101
will be described., Referring to
FIG. 2
, an NTSC format picture signal is supplied from an input terminal
100
to (m+1) multiplying devices
111
m
,
111
m−1
,
111
m−2
, . . . , and
111
0
. The multiplying devices
111
m
,
111
m−1
,
111
m−2
, . . . , and
111
0
multiply the received signal by respective coefficients and supply the calculated results to adding devices
112
m−1
,
112
m−2
, . . . , and
112
0
, respectively. Output signals of the adding devices
112
m−1
,
112
m−2
, . . . , and
112
0
are supplied to delay. registers
113
m−1
,
113
m−2
, . . . , and
113
0
, respectively. Output signals of the delay registers
113
m−1
,
113
m−2
, . . . , and
113
0
, are supplied to adding devices
112
m−2
,
112
m−3
, . . . , and
112
0
, respectively.
An output signal of the multiplying device
111
m
is supplied to a delay register
113
m
. An output signal of a delay register
113
m
is supplied to the adding device
112
m−1
. The delay registers
113
m−1
,
113
m−2
, . . . , and
113
0
delay their received signals by a delay time period T.
Thus, the NTSC format picture signal that is received through the input terminal
100
is supplied to the delay register
113
m
. The delay register
113
m
delays the NTSC format picture signal by a time period T. The resultant picture signal is supplied to the adding device
112
m−1
. The adding device
112
m−1
adds the output signal of the delay register
113
m
and the output signal of the multiplying device
113
m−1
and supplies the resultant signal to the delaying device
113
m−1
. The delay register
113
m−1
delays the output signal of the adding device
112
m−1
by the time period T and supplies the resultant signal to the adding device
112
m−2
. The adding device
112
m−2
adds the output signal of the delay register
113
m−2
and the output signal of the multiplying device
111
m−2
and supplies the resultant signal to the next delaying device.
Next, the similar process is repeatedly performed. The adding device
112
0
on the last stage adds the output signal of the delay register
113
0
and the output signal of the multiplying device
111
0
and supplies the resultant signal as the final output signal of the horizontal interpolation filter
101
(namely, the output picture signal of the horizontal interpolation process) to the vertical interpolation filter
102
through an output terminal
120
.
The structure of the vertical interpolation filter
102
is similar to that of the horizontal interpolation filter
101
. The vertical interpolation filter
101
performs a vertical interpolation process for the output signal of the horizontal interpolation filter
101
and supplies the resultant signal as a high-vision format picture signal to a high-vision receiver. In such a picture information converting process, an NTSC format picture signal can be displayed on a high-vision receiver.
In the above-described picture information converting process, an NTSC format picture signal is simply interpolated in the horizontal direction and the vertical direction. Thus, the resolution of the converted picture signal is the same as that of the original picture signal. In particular, when a normal moving picture is converted, the vertical interpolation process is performed as an intra-field process. In such a process, since the inter-field correlation of the picture is not used, due to a conversion loss, the resolution of the converted picture signal may deteriorate against that of the original picture signal.
To solve such a problem, the applicant of the present invention has proposed an apparatus that performs a class categorization adaptive process as a picture information converting process (see Japanese Patent Laid-Open Publication No. 6-205934). In the class categorization adaptive process, an input SD signal is categorized as a class corresponding to a three-dimensional (time-space) distribution of the signal level. Predictive coefficients pre-learnt for individual classes are stored in a memory. With the results of the class categorization and the predictive coefficients, a calculation corresponding to a predetermined predictive expression is performed so as to generate an optimum estimated value as an HD pixel.
In the class categorization adaptive process, with SD pixel data present in the vicinity of an HD pixel to be generated, the class categorization process is performed. Predictive coefficients are pre-leant for individual classes detected in the class categorization process. For a still picture portion, using the intra-frame correlation, an HD pixel value closer to a real value is obtained. For a moving picture portion, using the inter-field correlation, an HD pixel value closer to a real value is obtained.
Next, a real example of such a process for generating HD pixels y
1
and y
2
shown in
FIG. 3
will be described. The averages of frame differences of pixels present at the spatially same position are obtained for SD pixels m
1
to m
5
and SD pixels n
1
to n
5
. The obtained values are categorized as motion classes using predetermined threshold values. In addition, SD pixels k
1
to k
5
shown in
FIG. 4
are processed by ADRC (Adaptive Dynamic Range Coding) method. Thus, with a small number of bits, a class categorization that represents a spatial waveform can be performed.
For each class determined by the above-described two types of class categorizations, HD pixels y
1
and y
2
are generated by a calculation corresponding to the following linear expression (1).
y
=w
1
×x
1
+w
2
×x
2
+ . . . +wn×xn (1)
FIG. 5
shows an example of the arrangement of SD pixels x
1
, x
2
, . . . , and xn used in such a calculation. In this example, 17 SD pixels (n=17) are used. The predictive coefficients w
1
to wn used in formula (1) are pre-learnt. In such a process, since the class categorization that represents the amount of a motion and the class categorization that represents a spatial waveform are independently and adaptively preformed, a high conversion capabili
Kondo Tetsujiro
Uchida Masashi
Frommer William S.
Kostak Victor R.
Sony Corporation
LandOfFree
Picture conversion using field-by-field vertical inversion... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Picture conversion using field-by-field vertical inversion..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Picture conversion using field-by-field vertical inversion... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2942749