A Robust Image Transmission Scheme For Wireless Channels Based On Compressive Sensing

Dahua Gao,Danhua Liu,Youqian Feng,Qinli An,Fuping Yu
DOI: https://doi.org/10.1007/978-3-642-14932-0_42
2010-01-01
Abstract:Packet losses decrease the quality of image or video in multimedia applications. Robustness to packet losses is crucial to high quality image transmission in Internet or wireless channels where data loss often happens. Recently, compressive sensing (CS) has attracted great attention of both mathematicians and computer scientists, which shows that sparse signal or image can be recovered from far fewer measurements of it and the quality of CS-reconstructed image only depends on the number of measurements used in the reconstruction, not on which measurements are received. This fact will be helpful to resist packet loss. The aim of this paper is to present a robust still-image transmission coding scheme based on CS theory. If part of the bit-stream has been lost in the transmission, the original image can still be completely reconstructed from the remanent part of bit-stream (measurements). Numerical experiments show that high-quality image transmission is possible even with 30% random packet loss when this method is used for image transmission on wireless network. Besides being error-free, our proposed method without error control coding has low complexity.
What problem does this paper attempt to address?