A Novel Data Hiding Method based on Edge Detection and 2k Correction with High Payload and High Visual Quality

Abstract

In this paper a novel edge adaptive data hiding method is proposed. The proposed edge adaptive steganography method is improved version of the Bai et al.'s [31] (A high payload steganographic algorithm based on edge detection) method. The main aim of the proposed method is to achieve high payload with high visual quality. This method consists of preprocessing, edge detection, classification of the secret data, data embedding and data extraction phases. In the preprocessing phase 6 and 7 least significant bits (LSB) elimination are applied on the cover image to get a better edge detection in the next phase. As edge detection methods Sobel, Canny, Laplacian Of Gaussian (LOG), block-based edge detection and hybrid edge detectors are used. After the edge detection, secret data is divided into 2 classes and then these are embedded into edge pixels and texture pixels of the image. In order to increase the visual quality 2k correction is applied on the stego images. Modulus operator is utilized in the data extraction phase. In the experimental results, payload and visual quality measurements demonstrated the success of the proposed method.