Traditional and Deep Learning Techniques in Image Steganography: Recent Advances

For many decades however recently super increased; the importance of securing the private information of individuals or organizations. Therefore, numerous researchers are motivated to be involved in the challenge of building intelligent, flexible, secured, imperceptive, bumped bit embedding rates and robust models of Stenography. Image steganography is a way of concealing data, such as text, pictures, or videos, under an image that acts as a cover. Recent years have seen an increase in interest in deep learning technology, which has been proven to be a powerful mechanism in a number of uses, such as image steganography. This paper's primary goal is to investigate and clarify three main types of image steganography techniques: traditional, convolutional neural network-based, and generative adversarial network-based. An explanation of the datasets and evaluation measures is provided along with the approach. A comprehensive summary of various image steganography techniques is introduced. Finally, a discussion on the potential directions for this field's future is introduced.