A crisscross checking technique for tamper detection in halftone images

摘要

As the race for digital technological advancements accelerates, foremost is the troubling concern for better protection for digital data. Unequivocally detecting for tamper in data poses a great challenge to information security and image processing. In this paper, a crisscross checking technique is proposed for tamper detection in halftone images. First, a host image is divided into NxN blocks. The individual blocks will be scrambled with a random number into code-strings which will be concatenated into rows and columns code-streams. Finally, these code-streams will be applied to the MD5 and RSA algorithms to generate encrypted signed messages that will be hidden in the blocks. For the detection process, the row and column blocks will be checked for these signed messages which will be extracted, decrypted and compared to the rows and columns blocks of the image for dissimilar bits as sign of tamper. The tampered row and column blocks will be crisscrossed and the intersecting blocks will be identified as tampered areas. (c) 2004 Elsevier Inc. All rights reserved.