Adaptive Image Steganalysis Adaptive Image Segmentation using Enhanced Canny Edge Detection Algorithm

An expertise method to conceal secret messages in blameless mediums such as videos, audios, texts and digital images by cheating the Human Visual System (HVS) is termed as steganography. Blind Steganalysis is the striking technique to identify, excerpt hidden information from multimedia file with concealed secret information called stego media. To conceal the hidden message, adaptive image steganography determines the exact color adaptive regions or payload locations of the image. Adaptive image steganography's reverse in return is adaptive image steganalysis, in which the secret messages are dig out from the image's color adaptive regions for a familiar or unfamiliar steganographic algorithm that was used for steganography. In the first phase of the adaptive image steganalysis, the adaptive regions of the image are segmented in such a way that the color intensity values of the pixels are compared and grouped to form a region. Enhanced canny edge detection operator outperforms other segmentation algorithms. In the second phase, the statistical features of the pixels are extracted from the spatial domain region using Histogram of Oriented Gradient (HOG) and MatriX Quadtree (MX Quadtree) method. Feature selection from the extracted feature vector subset is done based on distance correlation coefficient method and Markov Random Field (MRF) Cliques method along with a ranking based wrapper approach. In the third phase, classification of