Distributed video coding with frame estimation at decoder

Abstract

In distributed video coding, input video stream is split into group of pictures of various lengths. Longer group of pictures is preferred as there is more temporal redundancy. However, the system time might increase as more frames need to be transmitted and be stored in the buffer. In this paper, we try to reduce the system time by proposing a model where only the odd-numbered Wyner-Ziv frames shall be transmitted from the encoder. At the decoder, the missing evennumbered Wyner-Ziv frames shall be estimated for full reconstruction of the video sequence. The simulation results show that the proposed model is more efficient as the output video sequence could be obtained in a shorter time. The estimated missing frames at the decoder are also of acceptable quality compared to the original frames at the encoder.