ROI based complexity reduction algorithm for H.264 encoder

Abstract

This paper proposes a region-of-interest (ROI) based H.264/AVC encoder which reduces computational complexity. A proposed fast ROI detection algorithm which can detect face-like regions as ROIs is applied to obtain accurate and small ROI so as to reduce the necessary coding effort of the encoder. The complexity reduction algorithm contains 3 methods: (1) the inter prediction mode selection based on quality difference and ROI status, (2) unequal coding efficiency degradation based on unequal bits allocation, and (3) the ROI boundary enhancement to reduce the coding complexity of center MBs of ROI. Experimental results show that proposed novel ROI detection and complexity reduction scheme can reduce 76.10% simulation time when QP difference between ROI and non-ROI is 20. Compared with previous work, 19.09% of encoding time is further reduced with similar performance degradation.