Toward Optimal Embedding Capacity for Permutation Steganography

Abstract

Bogomjakov et al. proposed a universal algorithm for permutation steganography. In this paper, we introduce a more effective algorithm. A theoretical analysis indicates that our algorithm achieves 99% of optimal when using 128 elements, while Bogomjakov et al.'s method attains only 98.34% of optimal even when employing as many as 1,048,576 elements. We conducted an experiment using a 3-D polygon dataset for comparison. The analysis and practical experiments demonstrate that our algorithm attains a capacity closer to the optimal but has the same runtime complexity as Bogomjakov et al.'s algorithm.