A single pass, high throughput reversible watermarking scheme for audio based on redundant embedding

Abstract

Reversible watermarking is a class of watermarking where we can recover the host (cover) signal losslessly along with distortion-free recovery of the watermark. The reversible watermarking algorithms proposed till date have been proposed for digital images. In this paper we propose a reversible watermarking scheme for audio signals. In general, reversible watermarking algorithms are computationally more expensive compared to their non-reversible counterparts. To reduce the run time latency of the proposed reversible audio watermarking scheme, a redundant watermark embedding technique is used which embeds the watermark without overflow checking of the watermarked audio sample values, but can reversibly retrieve the watermark utilizing the redundancy. The proposed technique also avoids the use of a location map for overflow checking, thus avoiding the need for multi pass, computation intensive data processing.