| 摘要 |
For large screen video displays using line-doubling to reduce scan line visibility, where a video source signal in standard interlaced format such as NTSC is converted to line-doubled non-interlaced format for progressively scanned display, visible motion degradations such as edge shimmer and strobe effects due to film-sourced video material are minimized by improved motion-detection video processing taught by this invention. In advance of actual display, successive fields are monitored for motion value in a first motion-detector and microprocessor comparator/analyzer where fields having high field-motion-value are detected as H-fields. Then, for display, each H-field is modified by pixel-averaging with a selected non-H-field. The regular and averaged fields are correctly sequenced to achieve vertical pixel alignment and smooth motion transitions thus minimizing film-source edge motion degradation. Film-mode entry and exit are accomplished automatically in a second motion-detector-analyzer that accomplishes fast film/video recognition and mode-switching. Film-mode entry and exit are each accomplished within a time period of one field, several times faster than in known art; such fast entry and exit are particularly important in avoiding visible motion defects under worst case conditions where regular video material is fragmented by frequent and/or non-synchronous spliced insertions of film-sourced or computer-generated portions.
|
