A robust content-based watermarking technique

Abstract

Geometric transformations change the image pixel positions where the watermark is embedded. As a result, a copyright verifier fails to detect the watermark even though the watermark is present in the transformed image. The contentbased watermarking schemes try to solve this problem by first dividing the image into many disjoint patches and then locating those patches with respect to the salient points of the image. Most of the existing content-based schemes are vulnerable to scaling and general affine transformations. They do not work well against high JPEG compression either. This paper presents a contentbased blind watermarking scheme using corners. The proposed scheme achieves robustness against affine transformations and signal processing attacks by using triangular patch normalization and spread spectrum watermarking technique respectively.

A Fast Corner Detector Based on the Chord-to-Point Distance Accumulation Technique

“…We believe such a fast and robust corner detector can be exploited in various applications including copyright protection [2], image matching [24], photogrammetry [4] and transformed image identification [25].…”

“…As a result, the distance from a point on the original curve to its location in the smoothed curve is high on and near a corner location but small if it does not have a corner nearby. Consequently, the maxima of the distance function 2 The big-oh notation 'O' denotes the asymptotic upper bound (the worst case running cost) of a function. See [19] for a rigorous definition.…”

“…Image features, for example, corners and interest-points, have applications in various areas such as computer vision [1], content-based copyright protection [2], transformed image identification [3] and photogrammetry [4].…”

A Performance Review of Recent Corner Detectors

“…Most of the applications (for example, for identification of appropriate image patches in feature-point based image copyright protection schemes [14]) require high repeatability and low localization error of the detected corners. There are also other applications (for example, extraction of building roof lines [1]) that require high repeatability, but may allow some localization error.…”

A steganalysis method in the DCT domain

“…They observed that the hidden information, the information hidden in the cover images, of stego-images are clustered in a plane while all other information of cover images are scattered more evenly in the whole space and have no other clusters. Awrangjeb and Lu [1] proposed micro and macro calibration methods that detect hidden information by calibrating the local and global distribution of the DCT coefficients of the image. All these methods employ high-dimensional feature vectors to describe the difference between cover and covert images, thus significantly affecting their performance in engineering applications.…”