Purpose
An online social media (OSM) is a powerful virtual network that facilitates global forwarding/sharing of messages, planning, analysis, and/or voting. Due to the rapid enhancement of accessibility of internet services, one may simply forward/share web content through OSM. These may include renowned OSM platforms Twitter, WhatsApp, Instagram, and Facebook to name a few. Such a practice of sharing web content without validating the authenticity of the source may have major political, social, or economic ramifications for society. The proposed research work aimed to propose a novel watermarking approach to reveal the first user/source of shared web content (image) on OSM.
Method
To authenticate the source, the combination of 10 digit mobile number, social security number (AADHAAR number in India), GPS coordinates, and specific code of the messenger app are used as a watermark. Prior to integration, the hamming code is utilized to encode the watermark, to make an approach more robust. In the embedding phase, the cover image is initially split into non-overlapping uniform blocks. Afterward, each block is subjected to Slantlet transformation (SLT). Moreover, four copies of the source-centric data are inserted during the watermark insertion process to achieve high reliability. The proposed method has been validated for effectiveness experimentally and compared with other closely related studies.
Results
The results revealed a higher level of robustness with a significant level of imperceptibility in terms of BER and PSNR respectively under various signal-processing attacks. In addition, the approach is determined to be fast enough for practical usage. Hence, the identification of the source of the shared content has been achieved to a higher degree.
Conclusion
A comparison with various existing approaches shows the applicability of the proposed methodology in terms of robustness, durability, and time complexity. The scope of the research will be broadened in the near future to advance in watermarking employing host images of varying sizes, attacks involving rotation and translation, and blockchain technology.