Various techniques for low bit rate video coding

Abstract

In this paper various techniques for low bit rate video coding are studied and implemented. The selection of an optimal regularshaped pattern set for very low bit-rate video coding, focusing on moving regions has been the objective of much recent research in order to try and improve bit-rate efficiency. An approach to Variable Size Block Matching is also implemented, based on the binary partitioning of blocks. If a particular block does not allow for accurate motion compensation, then it is split into two using the horizontal or vertical line that achieves the maximum reduction in motion compensation error. This method causes partitioning to occur along motion boundaries, thus substantially reducing blocking artifacts. In addition, small blocks are placed in regions of complex motion, while large blocks cover regions of uniform motion. Also the paper is a review of the block matching algorithms used for motion estimation in video compression. It implements and compares 7 different types of block matching algorithms that range from the very basic Exhaustive Search to the recent fast adaptive algorithms like Adaptive Rood Pattern Search.

Inpainting de imagens digitais por estimação de curvas de decaimento dos coeficientes wavelet via regularidade e espaços de Besov.

“…Uma forma de se realizar a detecção de erro se faz pela adição de uma informação de cabeçalho. Por exemplo, em uma transmissão de vídeo realizada em pacotes, cada pacote deve conter um cabeçalho e um campo de informação útil [64]. O cabeçalho contém um campo com o número sequencial do pacote, que pode ser utilizado no decodificador para detecção de perdas de pacotes.…”