An efficient Moving object, Encryption, Compression and Interpolation technique for video steganography

Video steganography approach has been widely used to preserve the secret information while transmitting through the internet. This approach has become attractive among many researchers due to its feature high Embedding Capacity (EC). In recent years, numerous methods are developed to secure the data. However, these methods still face problem in the area of imperceptibly and finding less distortion regions that is suitable for data hiding. To overcome this issue, a novel edge based Moving Objects (MO), compression technology, encryption method and interpolation data hiding technique based on moving regions detection that achieves less distortion is proposed in this paper. In the proposed system, first the video gets separated into frames, and edge based MO detection technique with graph system is used to identify the missing vertices. Then, compression technology with Rank Approximation (RA) and Prediction Error (PE) is applied for every frame. Next, we propose encryption scheme for MO which combines threshold and Pixel Value Differencing (PVD) to shuffle pixels, and non-MO applies circular shift encryption that preserve the pixels suitable for data hiding. Finally, the frames are up-scaled using weight based interpolation technique and data is hidden with various range inside Red Green Blue (RGB) channel by finding rough and smooth pixels. To ensure reversibility, any one channel is kept as the indication pixel for each pixel to store MO pixels. Experimental results reveals that the proposed method has significant improvement in the areas of compression level, MO identification and Quality Parameter (QP) testing. Moreover, Peak Signal-to-Noise Ratio (PSNR), EC, Regular Singular (RS) and PVD histogram analysis attacks obtained on the proposed method performs better than the other state-of-the-art techniques.