Robust watermarking algorithm for digital image based on SIFT feature points

Abstract

In this paper, we present a robust watermarking algorithm for digital image using SIFT (Scale-invariant feature transform) feature points and realizes the blind detection. The algorithm uses SIFT feature points on the host image to decide the embedding location and capacity, and adjusts the size of the binary watermark with Neighbor interpolation, and embeds the watermark into DCT low-frequency coefficients of some qualified sub-blocks for the host image. The algorithm has the following characteristics: (1) the watermark is embedded in the DCT coefficient based on scale-invariant feature points, which can effectively resist common attacks, such as Gaussian noise-adding, salt and pepper noise-adding, cropping, Wiener filtering; (2) the watermarking detection does not require original digital image. The experiments show that the algorithm has good invisibility and strong robustness.