Television – Format conversion – Line doublers type
Reexamination Certificate
2000-01-26
2002-09-24
Kostak, Victor R. (Department: 2711)
Television
Format conversion
Line doublers type
C348S581000
Reexamination Certificate
active
06456329
ABSTRACT:
The present invention relates to de-interlacing of video signals.
A motion picture, whether produced from film or electronic video information comprises a sequence of frames of picture information projected sequentially, i.e. a display of a sequence of still images progressing at a controlled rate. Images are displayed at a sufficiently high rate so that the viewer doesn't perceive the flicker of the individual frame images, e.g., 60 frames per second.
Where technology, bandwidth or other limitation does not allow information to be transmitted at full rate, methods transmitting information at a lower rate that will not unacceptably degrade the picture are used. Historically, television signals have been encoded in a interlaced format to increase viewer perception of the information of the picture. For example, television signals transmitted in the NTSC format include 60 interlaced fields per second, with each successive two fields making up a complete frame of picture information, typically in a 4:3 horizontal to vertical aspect ratio. An interlaced format means that each frame of picture information is separated into two fields of picture information, with each field including half of the lines of picture information in the frame, as in the simplified illustrations of two successive interlaced fields in FIG.
1
. In NTSC format, for example, each frame is captured in 480 horizontal lines, with 240 lines transmitted in each field, i.e. the even numbered lines are transmitted in one field and the odd numbered lines are transmitted in the next successive field, and so forth. For purposes of illustration, fewer lines are shown in FIG.
1
.
Many newer video standards, including the ATSC standard for digital video, utilize interlaced pictures because of the advantage it offers in display quality. On the other hand, modem display devices such as computer monitors and certain high-definition televisions employ a progressive scan wherein images are displayed as a sequence of frames without interlacing of lines or fields. I.e. each frame contains all the horizontal lines of a picture. Thus, a progressive display displays 480 horizontal lines per frame at a rate of 60 frames per second, twice as many as an equivalent interlaced display.
To display video received in an interlaced image encoding format on a progressive scan display requires conversion of interlaced video signals into progressive video signals, i.e. creating a series of progressive images from interlaced images. Problems associated with such conversion at full image resolution (i.e. without loss of picture detail) are that the resulting progressive image requires twice as much information as is contained in the interlaced image, and that massive memories needed to store fields of video information (field memories) and frames of video information (frame memories). Also, converting video data from one format to the other is very computationally intensive and various interpolations and other combinations of video data must be undertaken, typically millions of times per field.
Apparatus performing such full resolution conversions is very complex and expensive. An additional problem arises due to movement in the image between fields which when processed by the conversion computations often produces objectionable artifacts, i.e. objects appearing in two places or smeared or disturbing patterns. Conventional solutions tend to increase the complexity of the conversion process, and so increase the cost thereof in attempting to overcome these problems.
Even where so-called “simplified” de-interlacing schemes have been proposed, as in the example of the “bob” de-interlacing technique proposed by a major software company, the process remains computationally intensive. In the “bob” technique, the blank lines in each field are dropped, but then are in effect sought to be restored by a “times-two zoom” process to fill in the “missing” lines in an attempt to regain the vertical resolution and aspect ratio lost due to the dropping of one half the lines. This requires a computationally intensive process in which each pixel in the “missing” line is calculated from the pixel values of the two lines adjacent thereto. In addition, the full pixel content of each horizontal line is retained, thereby requiring at least a full field memory including memory storage for the full pixel content of each horizontal line, or a full frame memory if high-speed real-time interpolation computation is not provided.
Accordingly, there is a need for a simplified method for converting interlaced image fields into non-interlaced image frames. It is desirable that such method provide the advantage of simplified computation, even if the resolution of the resulting image frames is reduced.
To this end, the method of converting a sequence of interlaced image fields into a sequence of non-interlaced image frames of the present invention comprises:
removing each blank line of each image field, thereby to retain each line of each image field alternating with the blank lines therein, each such retained line including a sequence of image pixels; and
removing one-half the image pixels of the sequence of image pixels of each retained line of each image field.
In accordance with another aspect of the present invention, a storage medium encoded with machine-readable computer instructions for converting a sequence of interlaced image fields into a sequence of non-interlaced image frames comprises:
means for causing a computer to remove each blank line of each image field, thereby to retain each line of each image field alternating with the blank lines therein, each such retained line including a sequence of image pixels; and
means for causing the computer to remove one-half the image pixels of the sequence of image pixels of each retained line of each image field.
REFERENCES:
patent: 4603350 (1986-07-01), Arbeider et al.
patent: 5134480 (1992-07-01), Wang et al.
patent: 5305104 (1994-04-01), Jensen et al.
patent: 5473382 (1995-12-01), Nohmi et al.
patent: 5838385 (1998-11-01), Reder et al.
patent: 6166773 (2000-12-01), Greggain et al.
S. A. Booth, “Digital TV in the U.S.”, IEEE Spectrum, Mar. 1999, pp. 40-46.
A., Patti, “Digital Video Filtering for Standards Conversion and Resolution Enhancement”, (Ph.D. Thesis) abstract, Dec. 1995, 1 page.
“Microsoft Deinterlacing”,from http://www.microsoft.com/DDK/DDKdocs/win98ddk/dd-ddk 8h87.htm, date unknown, 2 pages.
E.B.Bellers and G. De Hann “Advanced Motion Estimation and Motion Compensated Deinterlacing”, International Workshop on HDTV '96 (paper No. H96-07), Abstract, Oct. 8-9, 1996, 1 page.
Reitmeier Glenn Arthur
Tinker Michael
Burke William J.
Kostak Victor R.
Sarnoff Corporation
LandOfFree
De-interlacing of video signals does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with De-interlacing of video signals, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and De-interlacing of video signals will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2872604