PVR system design of advanced video navigation reinforced with audible sound

This paper presents an advanced video navigation concept for Personal Video Recording (PVR), based on jointly using the primary image and a Picture-in-Picture (PiP) image, featuring combined rendering of normal-play video fragments with audio and fast-search video. The hindering loss of audio during conventional fast-search trick play is eliminated, by adding the sound associated to the normal-play video fragments. The normalplay fragments provide detailed information, whereas the fastsearch signal simultaneously presents a coarse overview, where it will be shown that the audio enhances the overall navigation efficiency. This system setup requires a specific DTV decoding process, combining multiple, independent audiovisual information signals. Experiments revealed that it is possible to decode all signals by efficient signal processing, thereby re-using the standard DTV decoding platform for decoding, both the normal-play audiovisual fragments and associated fast-search information signal. By applying a decoding concept for the fast-search navigation signal that allows exchanging the navigation refresh rate against execution cycles, it is ensured that real-time performance is obtained on an embedded CPU, as typically deployed in DVT platforms for PVR.