Reducing Spread Spectrum Image Steganography (SSIS) Extraction Errors With Feedback-Driven Adjustment

摘要

Steganographic techniques are useful to hide information in various types of common multimedia data for covert communication. Spread Spectrum Image Steganography (SSIS) is a data-hiding/hidden-communication method that uses digital imagery as a cover signal. This report examines the error sources in SSIS, their impact on payload throughput, and ways of minimizing these errors. We present a method that employs a feedback-driven adjustment to anticipate extraction errors and compensate for them without adversely affecting the detectability of the stegomessage. The technique, which is performed by the transmitter, does not require any additional effort on the part of the recipient We also describe an experiment using multiple cover images to evaluate the change in error rate as a function of varying input values along with modifications to the SSIS message embedding process. In conclusion, we propose enhancements to the feedback process, enabling it to be used not only for error reduction, but also as a means of limiting the detectability of a stegomessage.