Image Authentication Using Image Hashing Based on Ring Partition with Corner Inclusion

Abstract

Due to the availability of advanced image editing tools, existing image hashing techniques for image content authentication needs to be enhanced. Some of the recent better existing techniques based on ring partition cannot explore the complete content information available in the image. It is observed that around 50% of the corner information of the image remains unexplored. This paper presents an image hashing technique that tries to make use of corner information to generate the hash, which is more robust and discriminative than the existing ring partition based techniques. The algorithm based on ring partition and invariant vector distance that generates hash using intensity channel has been used to prepare the baseline system. The baseline system is extended by including the corner information of the intensity channel to develop the proposed model. It also uses the central orientation of the original image to make the system rotation invariant. The receiver operating characteristics (ROC) shows that the proposed model outperform some of the state-of-the-art methods.