Compressed sensing is famous for its compression performances
over existing schemes in this field. Conventional researches aim at reaching the
larger compression ratio at the encoder, with acceptable quality of reconstructed
images at the decoder. This implies the error-free transmission between the
encoder and the decoder. Unlike existing researches which look for
compression performances, we apply compressed sensing to digital images for
robust transmission in this paper. For transmitting compressed sensing signals
over lossy channels, error propagation would be expected, and the ways to
apply some means of protection for compressed sensing signals would be much
required for guaranteed quality of reconstructed images. We propose to transmit
compressed sensing signals over multiple independent channels for robust
transmission. By introducing the correlations between the compressed sensing
signals from different channels, induced errors from the lossy channels can be
effectively alleviated. Simulation results have presented the reconstructed
image qualities, which depict the effectiveness for the protection of compressed
sensing signals.