Television – Format conversion – Line doublers type
Reexamination Certificate
1999-02-17
2001-06-05
Eisenzopf, Reinhard J. (Department: 2614)
Television
Format conversion
Line doublers type
C348S452000, C348S715000, C375S240120
Reexamination Certificate
active
06243140
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to video data processing and, more particularly, to reducing the amount of buffer memory required to perform a decoding and format conversion operation.
BACKGROUND OF THE INVENTION
MPEG2 is a standard which has been proposed for the digital encoding of television signals. MPEG2 allows for signals to be encoded on either an interlaced or progressive basis.
The term interlaced is used to refer to image data that is represented using, e.g., two alternating image fields. One field of an interlaced video frame normally corresponds to the odd lines of pixels in the frame with the other field corresponding to the even lines of the frame. During display, the lines of one field are scanned, e.g., output to a display device. The lines of the second field are then scanned, so that they are interlaced on the display device with the lines of the first field. In this manner, when interlaced images are displayed, odd and even lines of image data are updated on an alternating basis.
In the case of progressive image data, image data is displayed sequentially, e.g., starting at the top left corner of an image proceeding to the bottom right corner. Thus, in the progressive image case, the lines of an image are displayed or updated on a sequential basis without lines being skipped.
In a series of progressive images, the positioning of the horizontal display lines is consistent from image to image. Accordingly, each newly displayed progressive image normally will completely replace the previous image on a display assuming that the images are the same size. In the case of interlaced images, each frame includes two fields which correspond to spatially different, e.g., horizontal odd and even lines, of a display device. Accordingly, in the case of interlaced images, each field updates only a portion of the displayed image. Because fields of an interlaced image normally correspond to images at different times, merely combining fields 1 and 2 of an interlaced frame can cause blurring and other image distortions when motion is present. For this reason, conversion of interlaced images to progressive images normally involves some form of motion detection and the application of processing which is a function of detected motion.
Presently television images are usually encoded to be displayed as interlaced images. Unfortunately, most computers are designed to display progressively scanned images.
The ability to efficiently convert between interlaced and progressive image formats continues to increase in importance due, in part, to the ever increasing use of computers. Notably, when television scenes or other data represented as interlaced images are to be displayed on a computer, they normally first have to be converted into progressive image data.
High speed memory is normally used in video processing applications which convert between image formats. This is so that real time, or near real time, processing of video data can be achieved. While the cost of memory has dropped considerably in recent years, memory still remains a significant cost component of many image and video processing systems. This is because a relatively large amount of memory is normally required for video applications. In consumer applications and other applications where cost is a concern, it is desirable to minimize the amount of memory required to implement an image processing system or device.
A known MPEG2 decoder
100
followed by a conventional interlace to progressive (I-P) scan converter
110
is illustrated in FIG.
1
. The (I-P) scan converter
110
operates to convert interlaced image data into progressive image data. Switching between intra-field and inter-field interpolation is performed in the system of
FIG. 1
as a function of detected motion in the image represented by the video data being decoded.
A frame memory includes sufficient memory to store data representing an entire frame. Known MPEG2 decoders such as the decoder
100
normally use three frame memories, a first anchor frame memory
102
, a second anchor frame memory
104
and a B-frame buffer
106
to decode MPEG 2 data as shown in FIG.
1
. The conventional I-P converter
110
which follows the MEPG2 decoder in
FIG. 1
uses an additional frame memory
112
for interpolation and motion detection purposes. Accordingly, a conventional MPEG2 decoder
100
followed by an I-P converter
110
normally requires a total of four frame memories.
In order to reduce the cost of video systems which perform decoding and conversion operations, there is a need for methods and apparatus which allow for a reduction in the amount of memory required to implement such systems. It is desirable that any new methods and apparatus be suitable for implementation in computer systems as well as televisions sets, set top boxes and other video applications.
SUMMARY OF THE PRESENT INVENTION
The present invention reduces the total amount of memory required to perform decoding and scan conversion operations, as compared to known decoding and scan conversion devices. This desirable result is achieved without significantly affecting video quality.
The reduction in required memory achieved by the present invention is accomplished by having an I-P conversion circuit utilize the same frame memory used to decode the images upon which a conversion operation is being performed.
In an exemplary MPEG2 decoding embodiment, anchor frame memories for storing I and P frames, and a B-frame buffer are implemented in a common memory device. This memory device is used for both decoder and scan conversion purposes. During I-P conversion difference signals between a present and a previous or a present frame and a subsequent frame, are used for motion detection. Inter-field interpolation requires a field delayed signal. Notably, in accordance with the present invention, the same frame memory used for decoding is used to provide the field delayed signal used during I-P conversion.
Thus, the present invention utilizes data included in decoder frame memories to detect moving areas for purposes of the I-P conversion process thereby avoiding the need for a separate I-P conversion frame memory.
In a specific exemplary embodiment, one of three frames, which is nearest to a present frame, is referenced for calculating frame difference signals, e.g., for motion detection purposes. In such an embodiment, a preceding or subsequent frame in a decoded video sequence, or field included therein, may be used for motion detection purposes. This is in contrast to known interlaced to progressive conversion systems which use only preceding decoded video frame data for motion detection purposes. Using data from subsequent frames for motion detection purposes allows frames stored in decoder memory, e.g., for use as reference frames for motion compensated prediction purposes, to be used for motion detection during the I-P conversion process of the present invention. The use of subsequent frame data for motion detection purposes facilitates the elimination of a separate frame memory dedicated to I-P conversion motion detection.
In an exemplary embodiment, interpolation processing is executed in a frame, e.g., a pair of fields, on an inter-field basis. By limiting interpolation to within a frame, and by using the above discussed motion detection technique the need for a separate frame memory for interpolation purposes is eliminated and the existing decoder frame memories can be used for I-P conversion.
Using the above discussed memory saving techniques, I-P conversion can be performed in accordance with the present invention without requiring addition frame memories beyond the number already present in a conventional decoder.
REFERENCES:
patent: 4706121 (1987-11-01), Young
patent: 5428397 (1995-06-01), Lee
patent: 5495300 (1996-02-01), De Haan et al.
patent: 5508747 (1996-04-01), Lee
patent: 5614940 (1997-03-01), Cobbley et al.
patent: 5617565 (1997-04-01), Augenbraun et al.
patent: 5666293 (1997-09-01), Metz et al.
patent: 5777682 (1998-07-01), De Haan et al.
patent: 5796437 (199
Eisenzopf Reinhard J.
Hitachi America, Ltd
Straub Michael P.
Straub & Pokotylo
Yenke Brian P.
LandOfFree
Methods and apparatus for reducing the amount of buffer... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Methods and apparatus for reducing the amount of buffer..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Methods and apparatus for reducing the amount of buffer... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2522049