Combined compressive sampling and image watermarking

A procedure for image watermarking in the presence of compressive sampling is proposed. The randomly chosen measurements from image blocks are used to carry the watermark. The image reconstruction based on a set of watermarked measurements is performed using the total variation minimization algorithm, allowing good quality of recovered image. The watermark can be efficiently detected within the recovered image, but only if we know the random measurement matrix used for image acquisition. This provides secure watermarking scenario. The watermark detection is tested in the examples using various images.