Title :
A new edge-based complexity measure for intra-rate estimation in H.264/AVC
Author :
Wang, Chao ; Mou, Xuanqin
Author_Institution :
Inst. of Image Process. & Pattern Recognition, Xi´´an Jiaotong Univ., Xi´´an, China
Abstract :
An efficient I-frame R-Q model is still needed by the rate control in the latest H.264/AVC video coding standard. Some recent works introduced the gradient-based image complexity measure into the frame-layer intra rate estimation in H.264/AVC, but the issue at the macroblock layer was not discussed. We propose a new edge-based complexity measure for this bit rate estimation problem at both the frame-layer and the macroblock-layer. The LOG operator is used in the edge extraction and an adaptive threshold selection method is proposed. The experiments show that the proposed edge-based measure has a good linear relation with the intra frame bit rate.
Keywords :
edge detection; estimation theory; video coding; H.264-AVC video coding; I-frame R-Q model; adaptive threshold selection; bit rate estimation; edge extraction; edge-based complexity measure; frame-layer; gradient-based image complexity measure; intra-rate estimation; macroblock layer; macroblock-layer; Automatic voltage control; Bit rate; Complexity theory; Correlation; Estimation; Image coding; Image edge detection; H.264/AVC; bit rate estimation; edge detection; image complexity; intra-frame;
Conference_Titel :
Signal Processing (ICSP), 2010 IEEE 10th International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4244-5897-4
DOI :
10.1109/ICOSP.2010.5656048