Recommendation on the Strategy of Using the

The purpose of this recommendation is to update the Recommendation on Scanning Formats formulated by the RACE Image Communication Project Line 4 in January 1994. This update takes into account the studies performed in WP2 within the framework of the Scanning Formats Extension. This paper also discusses the strategy to follow in order to adopt a progressive format for the coding of digital television. The work performed within the framework of the RACE 2110 HAMLET Extension on Scanning Formats, leads to the following statements : 1. Interlaced format has been chosen in the early years of television considering it was one of the most interesting solution to achieve compression with regard to the available technology 1]. 2. The improved quality of today's television sources and displays make the viewers less tolerant of the defects of the interlaced format (i.e. interline icker, line crawling and eld aliasing), especially for large displays, close viewing distance and high brightness levels 1]. 3. Besides the typical interlaced defects, interlaced sources further suuer from a loss of vertical deenition, deliberately introduced in order to reduce interline icker 1]. 4. With the change from analog to digital television, drastic changes will occur in our television system. These changes may be seen as a unique opportunity to change formats, implying only minor costs compared to the overall budget involved in such operation 1]. 5. Leaving space for the future, it could be envisaged as a wise step not to degrade the image quality at the very beginning of the television process (i.e. inside the camera), chosing a lossy scanning format 1]. 6. The choice of a scanning format needs to be done with care in order to avoid backward and lateral compatibility problems that would become diicult to overcome in the future 1]. 7. Compared to interlace, progressive-scan signals are more suitable for digital compression in terms of high correlation for vertical and temporal directions, motion estimation and compensation 1].