Enhancing Video Steganography with Nature-Inspired Deep RNN for Pixel Optimization 🌿

NATURE INSPIRED ALGORITHM FOR PIXEL LOCATION OPTIMIZATION IN VIDEO STEGANOGRAPHY USING DEEP RNN
Shamal Salunkhe, VJTI, India / Surendra Bhosale, VJTI, India

(Presented at the International Conference on Engineering, Science and Technology (IConEST) which took place on October 21-24, 2021, Chicago, USA, https://www.2021.iconest.net/ organized by International Society for Technology, Education and Science ((ISTES)-http://www.istes.org)

The steganography is applied on text, audio, image and video files. The steganography is useful for safe and secure data transmission. Video steganography is used to preserve confidential information of security applications. To improve security of the message, pixels locations are optimized using nature inspired algorithm. As conventional algorithms have a low convergence rate a new algorithm is proposed. A New algorithm is developed by combining two model algorithms namely, Water wave optimization (WWO) and Earth worm optimization (EWO) and is renamed as proposed algorithm Water-Earth Worm Optimization (WEWO) algorithm. The frames are preprocessed and extracted using Discrete Cosine transform (DCT) and structured similarity index (SSIM), respectively, as regular processing. For pixel prediction, the fitness function is obtained from neighborhood entropies in proposed algorithm. In this, method secret message is embedded with two level decomposition of Wavelet Transform (WT). In the proposed work tested with ‘CAVIAR’ dataset. The Proposed WEWO-Deep RNN algorithm performance is tested with modular noises such as, salt and pepper, pepper and white Gaussian. The proposed method gives enhanced performance, which is seen with parameters such as Peak Signal to Noise Ratio (PSNR), Mean Square Error (MSE), and Correlation Coefficient (CC) which defines image quality indices.