Fast video encoding algorithm for efficient social media service

초록

The Semantic Web, or web 3.0, is driving the evolution of internet by enabling more interaction with users of multimedia data. ITU-T/VCEG and ISO/IEC MPEG have recently completed a new joint standardization activity on video coding, called High Efficiency Video Coding (HEVC). In order to achieve the promise of Semantic Web, HEVC needs to provide good video data quality to end user in real time. This new standard contains significant improvements in picture quality for high resolution video. Main challange in this standard is the required computational time. In this paper, a novel early skip detection technique to reduce the encoder time is proposed based on statistical modeling. The proposed method was implemented on HM 7.0 reference software and experimental results show that the technique provides superior results for time reduction with only a marginal quality loss with all fast options set to "ON".