A Novel Method for Secure Image Steganography

Abstract

Steganography is the science that involves communicating secret data in an appropriate multimedia carrier. The secret message is hidden in such a way that no significant degradation can be detected in the quality of the original image. In this paper, a new technique for embedding messages inside images is proposed. The pixels for message embedding are chosen such that the distortion introduced after embedding will be minimum. A distortion function is designed to calculate the cost of embedding for each pixel. The function evaluates the cost of changing an image element from directional residuals obtained using a wavelet filter bank. The intuition is to limit the embedding changes only to those parts of the cover that are difficult to model in multiple directions, avoiding smooth regions and clean edges. A technique that introduces less distortion to the carrier image will generally cause changes that are more difficult to detect, therefore providing more security.