A Discrete Rajan Transform-based robustness improvement encrypted watermark scheme backed by Support Vector Machine

Abstract

A key challenge in the watermarking methodology is the sturdiness against various kinds of attacks even while upholding adequate level of security. Several watermarking techniques were recently proposed by embedding robust watermark into digital images. In general, most of the algorithms focus only on limited attacks to determine the watermark robustness. This study demonstrates an improved watermarking scheme based on encryption of the watermark image by utilising Discrete Rajan Transform (DRT) which is backed by Support Vector Machine (SVM). The encrypting technique is worked out by splitting the actual image into four segments, followed by shifting every individual row in each segment and employing DRT to calculate coefficients of the new rows. SVM, as a classifier, backs the extraction of the watermark image process by enhancing the watermark image detection with maximum possible correlation. Various original images as well as watermark images were employed to evaluate the proposed scheme. Furthermore, the evaluation process for the performance is conducted by experimenting with various attacks as well with normal operation. The experimental results prove that the proposed scheme results in considerable enhancement of security and robustness, making it ideal for security needed applications.