Image Steganography System based on Hybrid Edge Detector

Abstract

In the field of image steganography, edge detection based implantation methods play vital rules in providing stronger security of hided data. In this arena, researcher applies a suitable edge detection method to detect edge pixels in an image. Those detected pixels then conceive secret message bits. A very recent trend is to employ multiple edge detection methods to increase edge pixels in an image and thus to enhance the embedding capacity. The uses of multiple edge detectors additionally boost up the data security. Like as the demand for embedding capacity, many applications need to have the modified image, i.e., stego image, with good quality. Indeed, when the message payload is low, it will not be a better idea to finds more local pixels for embedding that small payload. Rather, the image quality will look better, visually and statistically, if we could choose a part but sufficient pixels to implant bits. In this article, we propose an algorithm that uses multiple edge detection algorithms to find edge pixels separately and then selects pixels which are common to all edges. This way, the proposed method decreases the number of embeddable pixels and thus, increases the image quality. The experimental results provide promising output.