Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
2000-05-10
2002-09-17
Kelley, Chris (Department: 2813)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
Reexamination Certificate
active
06452972
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to the field of image display systems, more particularly to methods and systems for detecting motion in a video data sequence.
BACKGROUND OF THE INVENTION
Video image display systems create moving images by rapidly displaying a sequence of still images, or frames. Display systems must rapidly produce each new frame in order to create the impression of smooth motion in the video sequence. Each frame is formed by displaying an orthogonal array of picture elements, or pixels. During each frame, every pixel in the array is assigned its own brightness and color value. Digital systems typically represent a single pixel using three values, each value representing the intensity of one of three component colors.
Due to the high frame rate required to smoothly portray moving objects, display systems require a very high data throughput. Early television broadcast standards, such as NTSC, developed a technique called interlacing to reduce the throughput requirements. Interlaced video systems, such as NTSC, PAL, SECAM, and some HDTV standards, transmit each frame as two sub-frames or fields. Each of the two fields that form a frame contain alternate scan lines from the frame. The first field typically contains all of the odd scan lines while the second field contains all of the even scan lines. Because the display forms the two sequential fields so quickly, the viewer's eye integrates the sequential fields into a continuous moving display. While the two separate fields are visually integrated into a single frame, flicker is reduced by projecting the image fields sequentially.
Modern image display systems, such as not computer displays and some HDTV standards, are non-interlaced. Non-interlaced video systems are typically called progressively scanned, or simply proscan, since the lines that form each image are scanned sequentially from top to bottom instead of being divided into two fields. Proscan display systems must have a higher frame rate than interlaced systems in order to avoid visible image flicker. Because of the higher frame rate, proscan systems typically display more information and have a higher resolution than comparable interlaced systems with a lower frame rate.
Some modern image display systems with relatively high bandwidths convert interlaced video signals to proscan in order to improve the display quality. Additionally, some display devices, such as the digital micromirror device (DMD), utilize proscan data conversion to compensate for a lack of image persistence.
Proscan conversion can introduce errors, or artifacts, into an image depending on what the video sequence is displaying and how the proscan conversion is being performed. A simple form of proscan conversion simply adds the even lines from a frame to the odd lines of a frame. Although this form of proscan conversion is preferred for still images, it creates problems when displaying moving objects. The problems arise from the fact that the two fields in an image frame do not represent the image at the same point in time. The image data is created by scanning the original image twice, once for every odd line and a second time for every even line, therefore the even-line field represents data one-half of a frame period later than the data represented by the odd-line field. The proscan conversion described above, which creates current frame images by filling in missing lines with pixel data from the prior field, causes misalignment in moving images. This misalignment is most obvious along the edges of a moving object since the edges will appear jagged. The same effect occurs in the center of a moving object, but unless there is a lot of contrast within the object the artifacts are not as noticeable.
Alternative forms of proscan conversion, which eliminate the effects of motion, are line doubling and line averaging. Both line doubling and line averaging use data from adjacent pixels of the same field to fill in the missing lines of the current field. Line doubling simply displays each line from the present field twice, once in its proper position and once in place of the subsequent or preceding line from the next field. When the next field is received, the display again uses each line of image data twice, once in its proper position and once in place of the preceding or subsequent line from the previous field. Line-averaging systems create a new line of image data based on the average of the image data for the lines above and below the created line. Because both the line-doubling and line-averaging methods only use data from only one time sample, they avoid the problems associated with simply combining the two image fields. Line-doubling and line-averaging, however, reduce the effective resolution of the image since they use less information to generate each image.
In order to maintain the highest effective resolution while avoiding motion artifacts, proscan conversion systems should compensate for motion in the image data. Ideally, the contribution of adjacent pixels from the same video field and from the same pixel in adjacent video fields should depend on the amount of motion in the video sequence. Therefore an accurate motion detection system is needed to allow optimization of the proscan conversion process.
SUMMARY OF THE INVENTION
Objects and advantages will be obvious, and will in part appear hereinafter and will be accomplished by the present invention which provides an improved method and system for measuring motion in a video image and for performing a proscan conversion on interlaced video data based on the improved motion measurement.
According to a first embodiment of the improved method of measuring motion, a field-difference motion value is calculated for a missing pixel. This field-difference motion value is used to select a proscan algorithm that will accurately de-interlace the video data.
According to another embodiment of the improved method of measuring motion, a field-difference motion value is determined by calculating a first prior-field average value equal to the average of same-pixel prior-field image data and same-pixel prior-field two-rows-prior image data, and determining the absolute value of the difference between same-pixel prior-row image data and the first prior-field average value data.
According to yet another embodiment of the improved method of measuring motion, a motion value is determined by determining a first prior-field average value equal to the average of same-pixel prior-field image data and same-pixel prior-field two-rows-prior image data, determining a first field-difference motion value equal to the absolute value of the difference between same-pixel prior-row image data and the first prior-field average value data.
According to yet another embodiment of the improved method of measuring motion, a motion value is determined by determining a second prior-field average value equal to the average of same-pixel prior-field image data and same-pixel prior-field two-rows-latter image data, and determining second field-difference motion value equal to the absolute value of the difference between same-pixel prior-row image data and the second prior-field average value data.
According to yet another embodiment of the improved method of measuring motion, a first field-difference motion value is determined by determining a first prior-field average value equal to the average of same-pixel prior-field image data and same-pixel prior-field two-rows-prior image data, determining a first field-difference motion value equal to the absolute value of the difference between same-pixel prior-row image data and the first prior-field average value data, a second field-difference is determined by determining a second prior-field average value equal to the average of same-pixel prior-field image data and same-pixel prior-field two-rows-latter image data, and determining second field-difference motion value equal to the absolute value of the difference between same-pixel prior-row image data and the second prior-field average value data, and a
Brady III Wade James
Brill Charles A.
Kelley Chris
Philippe Gims
Telecky , Jr. Frederick J.
LandOfFree
Motion detection using field-difference measurements does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Motion detection using field-difference measurements, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Motion detection using field-difference measurements will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2904221