Adaptive Deblocking Filter Peter List, Anthony Joch, Jani Lainema, Gisle Bjøntegaard, and Marta Karczewicz IEEE TRANSACTION ON CIRCUIT AND SYSTEM JULY, 2003

Outline Introduction Boundary analysis Filtering Result Conclusion

Introduction Blocking artifacts Source of blocking artifacts discontinuities on the edges of the blocks Source of blocking artifacts block-based integer discrete cosine transforms (DCTs) motion compensated prediction

Introduction two approaches to integrate deblocking filters into video codec post filters only operate on the display buffer outside of the coding loop loop filters operate within the coding loop

Introduction Loop filtering has several advantages over post filtering guarantee a certain level of quality no need for an extra frame buffer in the decoder typically improve both objective and subjective quality

Boundary analysis error distribution in a 4 x 4 block coding errors are larger near the block boundaries than in the middle of the block 122 107 106 111 103 102 101 112 100 98 108 118 105 120

Boundary analysis H.264 deblocking filter is adaptive on several levels slice level block-edge level sample level

Boundary analysis block-edge level

Boundary analysis sample level

Boundary analysis slice level

Boundary analysis