High Performance Lossless Data Embedding Using a Moving Window

Abstract

This paper proposes a new lossless data embedding algorithm on spatial domain of digital images. A single key parameter is required to embed and extract data in the algorithm instead of embedding any additional information such as the location map. A  × window slides over the cover image by one pixel unit, and one bit can be embedded at each position of the window. So, the ideal embedding capacity equals to the number of pixels in an image. For further increase of embedding capacity, new weight parameters for the estimation of embedding target pixels have been used. As a result, significant increase in embedding capacity and better quality of the message-embedded image in high capacity embedding have been achieved. This algorithm is verified with simulations.