Distributed Video Coding Based on Selective Data Pruning and High-order Edge-directed Interpolation

In the current distributed video coding (DVC), a low resolution video sequence is generated with spatial and/or temporal downsampling at the encoder. At the decoder side, interpolation is performed and the interpolated pixels are further refined by using a error-correcting code such as Turbo codes or LDPC. In our previous work, we proposed a spatial domain DVC which uses a line-based downsampling method called selective data pruning (SDP). SDP is used for reducing the spatial frame size at the encoder. The obtained frames are then encoded and transmitted. Received frames are decoded and interpolated back to their original size by the single frame-based interpolation called the high-order edge-directed interpolation. However, artifacts are visible in the interpolated images because the positions of pruned lines are fixed for all GOP frames and the pixel continuity between frames are not sufficiently taken into account. In this paper, we focused on the image quality improvement of interpolated frames by using the multiframe-based interpolation. Experiments are performed to evaluate the quality of interpolated images and the rate distortion performance.