Key points are not available for this paper at this time.
Background: Image steganography is a technique that conceals confidential data within an image. The conventional technique of image steganography aims to effectively conceal confidential data within the host image, prioritising the secrecy of the payload over its capacity. However, it is necessary to enhance the quality of the steganography image to ensure optimal perception by the human visual system. Methods: The hidden image is subjected to the Discrete Wavelet transform for transformation, and subsequently encrypted using an advanced encryption technique (AES) to enhance the image's resistance to detection. The Deep Neural Network, consisting of Hiding and Extraction networks, enhances the steganography capacity by facilitating the concealment and retrieval of full-size images. Results: Some of the relevant metrics are evaluated and the outcome proves that the proposed method exhibits increased peak signal to noise ratio (PSNR), Structural similarity (SSIM) index, entropy, and reduced mean square error values than existing methods.
Hawkinson et al. (Fri,) studied this question.