|
Transform-domain Wyner-Ziv video coding for 1K-pixel visual sensors Host Publication: ACM/IEEE International Conference on Distributed Smart Cameras, ICDSC2013 Authors: F. Verbist, N. Deligiannis, W. Chen, P. Schelkens and A. Munteanu Publisher: IEEE Publication Date: Nov. 2013 Number of Pages: 6 ISBN: 978-1-4799-2166-9
Abstract: Efficient coding of extremely low-resolution video has received a mere shadow of the attention given to their high-definition counterpart. However, low-resolution video still proves suitable for video processing and may enable the design of advanced video-based applications in a cost-effective manner. Indeed, visual sensors with modest frame resolutions can be designed to be small, low-cost and low-power. This paper introduces a novel efficient codec, specifically designed to handle extreme low-resolution video captured by such visual sensors. By applying Wyner-Ziv coding principles, the encoding complexity is reduced since temporal redundancy is exploited at the decoder. The proposed system outperforms H.264/AVC Intra and is therefore particularly attractive to wireless, resourceconstraint low-resolution visual sensor applications by reducing the amount of transmitted data at a reduced computational cost.
|
