Multimedia data generation algorithm for multi-user video conferencing system over ad-hoc networks

Abstract

In this paper, a multimedia data generation algorithm for multi-user video conferencing system over ad-hoc networks is proposed. In the proposed method, the scene-changed frame is estimated using the number of bits pre-encoded by the hardware accelerator and the percentage of the intra-coded macroblocks in each frame. The resulting scene-changed frames are re-encoded as the intra-coded frames and then transmitted to the receiver. Compared to the bit-stream generated from encoding the entire sequence, the proposed algorithm achieves up to about 50% reduction of the bit-rate without skipping the key frames while maintaining the perceptual video quality.