Issues with digital watermarking and perceptual hashing

Robust identification of audio, still images and video is currently almost always associated with watermarking. Although being a powerful tool, there are some relevant issues with the use of watermarking. In this paper we review these issues, and at the same time propose to reconsider the older technique of robust feature recognition as a serious alternative. Moreover, we argue that not only in the context of content recognition, but also for other applications, a benefit is to be expected from the combination of robust feature recognition and digital watermarking.