Blind Compressive Sensing Detection of Watermark Coded by Limited-Random Sequence

Chao Zhang,Jialuo Xiao,Yaxin Zhang
DOI: https://doi.org/10.1587/transfun.e98.a.1747
2015-01-01
Abstract:Due to the fact that natural images are approximately sparse in Discrete Cosine Transform (DCT) or wavelet basis, the Compressive Sensing (CS) can be employed to decode both the host image and watermark with zero error, despite not knowing the host image. In this paper, Limited-Random Sequence (LRS) matrix is utilized to implement the blind CS detection, which benefits from zero error and lower complexity. The performance in Bit Error Rate (BER) and error-free detection probability confirms the validity and efficiency of the proposed scheme.
What problem does this paper attempt to address?