International Scholarly Research Network ISRN Signal Processing Volume 2012, Article ID 805346, 6 pages doi:10.5402/2012/805346 Research Article Sample-Level Filtering Order for High-Throughput and Memory-Aware H.264 Deblocking Filter Guilherme Correa, 1 Luciano Agostini, 2 and Luis A. da Silva Cruz 1 1 Instituto de Telecomunicac ¸˜ oes (IT) University of Coimbra—FCTUC, Polo II Universidade de Coimbra, Pinhal de Marrocos, 3030-290 Coimbra, Portugal 2 Group of Architectures and Integrated Circuits (GACI) Federal University of Pelotas, UFPel Campus Universit´ ario s/n, P.O. Box. 354, 96010-900 Pelotas, RS, Brazil Correspondence should be addressed to Guilherme Correa, guilherme.correa@co.it.pt Received 5 March 2012; Accepted 10 April 2012 Academic Editors: E. Ciaccio and P. Yin Copyright © 2012 Guilherme Correa et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. This paper presents a new sample-level filtering order for the Deblocking Filter process of the H.264/AVC video coding standard to be used instead of the traditional block-level order presented in previous works. This processing order allows a better exploration of the parallelism in the filtering process by reducing data dependencies in comparison to other works. The proposed sample- level order allows four parallel and independent samples filtering simultaneously, completing one complete macroblock filtering in fewer cycles and requiring less memory space than the related works. The proposed filtering order can be applied to the Deblocking Filter presented in a conventional H.264/AVC encoder or decoder and to the H.264/SVC interlayer Deblocking Filter. When compared to the original H.264/AVC filter and to the best related work found in the literature, the proposed scheme achieves a reduction of 72% and 25% in the number of clock cycles and a memory usage decrease of 75% and 43%, respectively. 1. Introduction The quality of a video may be treated in both objective and subjective fields. In the first case, distortion metrics are used in order quantify the quality of the video through the calculation of the signal-to-noise ratio between the original video signal and the final video signal. The most widely used distortion metric is called PSNR (peak signal-to-noise ratio). On the other hand, the subjective quality of a video is much more difficult to be tested or measured. Since it depends on how the image is perceived by a viewer, its measurement is more expensive in terms of time and human resources. Several video coding standards provide tools that deal with both objective and subjective video quality. However, the H.264/AVC standard [1] defines as mandatory the use of a module responsible for increasing the image subjective quality. This module, called Deblocking Filter [2], increases the image subjective quality through smoothing operations over the pixels. These operations are performed in order to decrease the perception of some block-shaped artifacts introduced in the frame by a high quantization level. This is an adaptive filter, since it can distinguish between real borders of an image and the borders introduced by high quantization steps. The Deblocking Filter of H.264/AVC is placed at the end of the coding process, as shown in Figure 1, where ME, MC, T, Q, T -1 , and Q -1 stand for Motion Estimation, Motion Compensation, Transform, Quantization, Inverse Quantiza- tion, and Inverse Transform, respectively. The information coded through interframe or intraframe prediction is added to the residuals and the resulting content is then filtered by the Deblocking Filter. After that, the image can be stored in order to be used during the coding of the next frame (in case of interframe coding) or during the coding of the next blocks of the same image (in case of intraframe coding). The H.264/AVC standard supports scalability through its Annex G, called Scalable Video Coding (SVC) extension [3], whose operating principle defines a hierarchy of layers where a base layer represents the video signal at a given minimum temporal, spatial, and/or quality resolution, and additional