Abstract

In 2012, Lee et al. proposed an interpolation enlargement technique using neighboring pixels as the base on which secret messages are embedded in enlarged interpolated pixels. This method can effectively predict the pixel between two adjacent pixels so as to reduce the distortion caused by the hidden secret messages. However, in this study we found that while Lee et al.’s neighboring pixel prediction method can successfully achieve a predictive value, embedded secret messages of different sizes will cause great distortion when a large secret message is hidden in the predicted value. Therefore, the purpose of our research is to determine how to process secret messages in advance to reduce the harm caused by embedding. We propose a secret message reduction strategy that changes positive secret messages to both positive and negative to decrease their values and thereby reduce stego-image distortion.