Image Steganography Based on Edge Detection Algorithm
Ismail Kich, El Bachir Ameur, Youssef Taouil · 2018
This article proposes a new steganography technique to embed secret message in grayscale images. It is an adaptive edge scheme that dissimulates the secret data in the edge pixels. The selection of the edge pixels is based on oversegmentation using Modified Simple Linear Iterative Clustering (M-SLIC). The M-SLIC algorithm makes it possible to segment an image into K regions called superpixel. This segmentation is very useful for the detection of objects and the contours between these objects. The number of superpixel K depends on the amount of data to be embedded and the regularity of the cover image. The experiment results have shown that the proposed technique improve the performance of the stego image in terms of capacity, imperceptibility and robustness compared to recent steganographic techniques.