High Capacity Data Hiding in Binary Document Images

In this paper, we propose a high capacity data hiding method in binary document images towards semi-fragile authentication. Achieving high capacity in binary images with strict imperceptibility criterion is found to be a difficult task. In this method, noise type pixels are selected for pixel-wise data embedding using a secret key. The data hiding process through pixel flipping introduces some background noise in watermarked images and could preserve relevant information. The reversible nature of noise pixel patterns used in flipping process enables blind detection and provides high watermark capacity illustrated in different test images. After extraction process, the background noise is removed to generate the noise-free version of the watermarked image.