Efficient pre- and post-processing schemes for block-based low bit-rate video coding

Abstract

This study aims at finding efficient pre- and post-processing methods for low bit-rate video transmission using a conventional block-based video coder. As a post-processing scheme, this paper proposes a method to remove blocking and ringing artifacts in low bit-rate block-based video coding. The proposed deblocking filter has 2 separate filtering modes, which are selected by pixel behavior around the block boundary. In each mode, proper one-dimensional filtering operations are performed across the block boundary along horizontal and vertical directions, respectively. Even though the proposed filtering is quite simple, it improves both subjective and objective image qualities for various image features. The modular structure of the proposed algorithm makes it possible to adjust the overall complexity depending on the available computational power of the decoder. In addition to the proposed post-processing scheme, a pre-filtering scheme is also proposed for noisy video sequences. Since pre-filtering removes camera noise and improves coding efficiency dramatically, its efficient implementation has been an important issue. Based on the approximated generalized Wiener filtering and two-dimensional DCT factorization, this paper introduces an efficient pre-filtering scheme performed inside a video encoder. For intra and inter block coding, pre-filtering is applied for original image blocks and motion-compensated error blocks, respectively. Since the proposed pre-filtering operation is actually performed by scaling transformed coefficients, the computational complexity is marginally increased and the overall architecture of a conventional video encoder is maintained. Notwithstanding its simplicity, the proposed pre-filtering scheme gives good filtering and coding performance for noisy video sequences.