Television – Format conversion – Field rate type flicker compensating
Reexamination Certificate
2000-12-21
2003-11-11
Miller, John (Department: 2611)
Television
Format conversion
Field rate type flicker compensating
C348S457000, C348S449000, C348S558000
Reexamination Certificate
active
06646684
ABSTRACT:
TECHNICAL FIELD
The present invention relates to a picture converting apparatus and a method thereof suitable for use with a television receiver, a VTR, and so forth.
RELATED ART
Conventionally, as television scanning field frequencies, 50 Hz (for PAL system and SECAM system) or 60 Hz (for NTSC system) have been widely used. When the field frequency is relatively low—for example 50 Hz, a so-called flicker of which a large area of the display screen flickers takes place.
To suppress the field flicker, an apparatus that doubles the number of fields of an input picture signal is known. Such a process is referred to as field double speed process.
FIG. 17
shows one example and another example of the pixel structure obtained by the field double speed process. In
FIG. 17
, the vertical axis represents vertical position V, whereas the horizontal axis represents time t. As the result of the field double speed process, an output picture signal having a 1/2 field period of an input picture signal is formed. Since the input picture signal is an interlaced signal, the vertical position (line position) of an input pixel (denoted by a white circle) of one field deviates by 1/2 between from that of the chronologically adjacent field.
The pixel structure shown in
FIG. 17A
is obtained by performing a double speed process in such a manner that double speed fields of an output picture signal are interlaced. Conventionally, to form output pixels of a field that are not present in an input picture signal, three pixels of the input picture signal are supplied to a median filter. In
FIG. 17A
, an output pixel (denoted by a star mark) is generated by the median filter with three input pixels placed in a triangular area denoted by dotted lines. The median filter outputs a pixel value that has the median pixel values of the three input pixels.
FIG. 17A
shows the process of the median filter for a part of output pixels. The median filter generates all output pixels other than those that match the positions of the input pixels. An output picture signal having the pixel structure shown in
FIG. 17A
is referred to as ABAB type.
On the other hand, in the pixel structure of an output picture signal shown in
FIG. 17B
, a double speed process is performed in such a manner that pixels of a new field are present at the same vertical positions as those of the chronologically early field of the input picture signal. In
FIG. 17B
, by repeatedly outputting the chronologically early field, the double speed process can be performed. An output picture signal having the pixel structure shown in
FIG. 17B
is referred to as AABB type.
In the AABB type, since the same field is repeatedly output, the double speed process does not improve the resolution in the chronological direction. Instead, when a picture is panned by a video camera, a dual picture takes place. Thus, the resultant picture quality may deteriorate. The ABAB type is advantageous against the AABB type with respect to the chronological and spatial resolutions (resolutions in the chronological direction and in the spatial resolution). However, in the ABAB type, three input pixels contained in chronologically early and late fields are processed by a median filter so that an output pixel is interpolated. Thus, when all the screen moves (for example, characters are horizontally scrolled), a vertical line looks like a comb. For example, when numeric characters and alphabetic characters are horizontally scrolled (such as stock information), their visibility deteriorates.
DISCLOSURE OF THE INVENTION
Therefore, an object of the present invention is to provide a picture converting apparatus and a method thereof that allow low chronological and spatial resolutions as a problem of the AABB type to be solved.
Another object of the present invention is to provide a picture converting apparatus and a method thereof that allow the ABAB type or the AABB type to be selected corresponding to the pattern of an input picture so as to impose the picture quality thereof.
To solve the above-described problem, the invention of claim 1 is a picture converting apparatus for converting an input picture signal into an output picture signal, the pixel position of a first field of the input picture signal being different from the pixel position of a second field thereof, the first field being adjacent to the second field, the field frequency of the output signal being N times higher than the field frequency of the input picture signal, where N is an integer that is 2 or larger, comprising:
a class deciding portion for designating each of all fields of the output picture signal that are not present in at least the input picture signal to a considered field and deciding a class for each considered pixel of each considered field corresponding to a plurality of pixels of the input picture signal, the plurality of pixels being decided corresponding to each considered pixel of the considered field,
a memory portion for storing pre-obtained predictive information,
a predictive pixel selecting portion for selecting a plurality of pixels of the input picture signal for each considered pixel, and
a pixel generating portion for generating each considered pixel of the output picture signal corresponding to predictive information for a class decided by the class deciding portion and a plurality of pixels selected by the predictive pixel selecting portion,
wherein the pixel position of the output picture signal varies at intervals of fields equal to N, and
wherein the pixel position of each field matches the pixel position of the first field or the second field of the input picture signal.
In addition, the invention of claim 13 is a picture converting method for converting an input picture signal into an output picture signal, the pixel position of a first field of the input picture signal being different from the pixel position of a second field thereof, the first field being adjacent to the second field, the field frequency of the output signal being N times higher than the field frequency of the input picture signal, where N is an integer that is 2 or larger, comprising the steps of:
designating each of all fields of the output picture signal that are not present in at least the input picture signal to a considered field and deciding a class for each considered pixel of each considered field corresponding to a plurality of pixels of the input picture signal, the plurality of pixels being decided corresponding to each considered pixel of the considered field,
selecting a plurality of pixels of the input picture signal for each considered pixel, and
generating each considered pixel of the output picture signal corresponding to predictive information for a class decided at the class deciding step and a plurality of pixels selected at the predictive pixel selecting step,
wherein the pixel position of the output picture signal varies at intervals of fields equal to N, and
wherein the pixel position of each field matches the pixel position of the first field or the second field of the input picture signal.
According to the present invention, when an output picture signal of AABB type is generated, a pixel value of an output field that is not present in an input picture signal is generated by a class categorizing adaptive process with picture information of input fields that are immediately early and late input fields thereof, in comparison with a conventional method for simply repeatedly outputting the same field, a field that is output second time has picture information corresponding to the output time. Thus, a problem of the AABB type that does not have a chronological resolution can be solved. In the class categorizing adaptive process, a class is detected with a plurality of pixels of an input picture signal. With an estimating predicting expression that is optimum for the detected class, a pixel value of an output picture signal is generated. Thus, in comparison with the case that a pixel is interpolated with a median filter, the resolution of the output picture signal can be improved.
In
Hattori Masaaki
Kondo Tetsujiro
Tatehira Yasushi
Frommer William S.
Frommer & Lawrence & Haug LLP
Miller John
Sony Corporation
Yenke Brian
LandOfFree
Image conversion device 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 Image conversion device and method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image conversion device and method will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3140596