Improved image watermarking using guided image filtering

Abstract

An improved version of image watermarking method based on regularized filter is proposed in this paper. The proposed method applies the concept of guided image filtering to improve the watermark extraction performance, while preserving the property of blind watermark detection. In the embedding process, the S component of the HSV color space is selected to carry a binary watermark image. In the extraction process, a guided image filter is introduced and utilized to estimate the original embedding component. The result is subtracted with the watermarked component to acquire the extracted watermark without using the original un-watermarked component. At equivalent quality of watermarked images in terms of weighted Peak Signal to Noise Ratio, the results demonstrate that our proposed method is capable of providing a higher quality of extracted watermark in terms of watermark to noise ratio. In addition, it provides higher robustness against various types of attacks from Stirmark, compared to the previous method.