Information Security by Hiding Data in Binary Images based on Block-Diagonal Partition Pattern

Abstract

Due to the shift from paper to electronic paradigm, much information is transferred in the form of binary images in multiple domains. This includes medical imaging, military communication, business transactions, and banking. During transmission, the confidentiality and security of information is compromised. Steganography and watermarking are the most popular means to serve this purpose. In both the techniques, secret data and a watermark are embedded respectively in cover images taking care that it is imperceptible. Hiding data and keeping changes caused by hiding unnoticeable is a significant challenge for binary images. This paper proposes a block diagonal partition pattern based method for hiding data in binary images. Image blocks of size 3X3 are considered to apply this method. Based on the number of black and white pixels in each diagonal partition, the potential blocks to carry data bits are identified. This is before and after embedding. The connectivity among pixels in each partition is also checked to declare block's suitability to carry data bit which takes care of distortion factor. The central pixel of a block chosen as a data bit carrier. For extracting the hidden data, the original image does not required as all blocks holding the data satisfy the embeddability criteria before and after hiding. The technique encrypts data twice to keep data more secure. The result shows that our method is secure and has good hiding capacity causing less distortion.