Content-based image watermarking via public-key cryptosystems

Digital watermarking is a technique to insert an information-carrying digital signature into a digital media so that the signature can be extracted for variety of purposes including ownership authentication and content verification. We examine the weaknesses against common watermarking attacks of blockwise independent and content-based watermarking algorithms for image integrity verification, and implement a new and more secure invisible fragile public-key watermarking algorithm for color or grayscale images that increases the message digest size from the proposed 64 to 128 bits using the same small-size blocks and maintaining high-quality watermarked images and accurate localization of image changes. Our watermarking technique is capable to detect any changes made to the image since the time it was stamped, any changes to the pixel values and also to the dimensions of the image will be automatically detected and localized. Our scheme consists of a watermark-insertion process that uses a private key to embed a watermark image into a cover image, and a watermark-extraction process that uses a public key to extract the watermark from the watermarked image. The embedded watermark can only be extracted by someone who has possession of a proper verification key.