Adaptive Steganography Based on Syndrome-Trellis Codes and Local Complexity

Syndrome-trellis codes provide a flexible framework for adaptive steganography, but how to construct the proper distortion weight function is still a disturbing question. In this paper, we propose to use an exponential function based on the local complexity to define the risk caused by the data embedding. And the steganography based on the payload-adaptive embedding risk and syndrome-trellis codes is presented. Finally, an experiment is performed to verify its validity.