Method, software, and device for hiding data in binary image, while preserving image quality

Abstract

A method of embedding watermark data into a two-colour (binary) image includes dividing the image into blocks and assessing the suitability of each block to embed a bit of watermark data by assessing whether or not the flipping of a defined pixel in each block affects the visual attributes of said block in manner to be perceptible by the human eye. Data is only embedded in those blocks determined to be suitable for data embedding, by flipping the defined pixel, as required. A recipient of the document may similarly assess which blocks contain watermark data, by assessing the suitability of each block in the document to embed such data. Conveniently, watermark data may be extracted without further information about the data’s location within a document.

Haiping Lu

Professor of Machine Learning, Head of AI Research Engineering, and Turing Academic Lead

I am a Professor of Machine Learning. I develop translational AI technologies for better analysing multimodal data in healthcare and beyond.