An edge-based image steganography method using modulus-3 strategy and comparative analysis

Abstract

Steganography or “Covered Writing” is a security tool for hiding secret information inside any media or object of interest. It provides secure communication between the sender and the receiver. The spatial domain image steganography techniques such as the Least Significant Bit (LSB) substitution and edge-based hiding schemes have been vastly exploited in the literature. An improved LSB method using modulus-3 strategy can be found in the literature, which has the advantages of high embedding capacity and resistance against Subtractive Pixel Adjacency Matrix (SPAM) steganalysis. However, one drawback is that it hides data sequentially. The edge-based hiding techniques have the advantage that it can tolerate significant variations in edge pixels and fulfill all the requirements of the Human Vision System (HVS), but it has limited in embedding capacity. The proposed work is a reversible data hiding scheme that takes the advantages of both the edge-based algorithm and the novel modulus-3 hiding scheme. At first, we converted the message into ternary data, and then we applied the novel modulus-3 strategy to hide two ternary bits in the edge pixel of the image thereby taking the advantages of both modulus-3 strategy and edge-based hiding approach. We have implemented four edge detection algorithms like Laplacian, Canny, Sobel, and Prewitt to find four different edge areas from the cover image. For each edge area, we applied the modulus-3 strategy to hide the entire ternary message, and then, we made a comparative analysis based on the performance of four edge detectors, and finally, we concluded.