This work's primary goal is to secure the transmission of text hidden within the cover image using steganography over a public network of computers. Steganography is a powerful tool for concealing information within a cover image so that the concealed message remains undetectable. As a result, steganography refers to concealed writing. The secure transmission of information over a public network communication channel using steganography occurs in two stages, the first on the sender side and the second on the receiver side. In the first phase, steganography is normally applied to conceal the encrypted information within the image as a cover. The encrypted data is implanted inside the cover image using an improved least significant bit steganography method. The secret key obtained by the embedding algorithm is shared with the message retrieval algorithm on the receiver side to retrieve the message. Finally, the embedded text message is identified using a hybrid convolution regression adaptive integrated neural network (CRAINN) approach.
Top1. Introduction
Improving data communication mechanism and the widespread use of multimedia technology has resulted in rapid advances in everyday life (Abdel Wahab, O. F. et. al., 2019). The process of protecting confidential information from unauthorized access has been practiced since ancient times. There were two methods for dealing with such a problem: cryptography and steganography (Zaynalov, N. R. et al., 2019).
Cryptography is defined as a process of comprehensive information security that involves encoding messages at the sender end to make them incomprehensible. In cryptography, data is safely transmitted on a public network of computers by using cryptographic techniques that make it much more difficult for intruders to access classified non-public information (Rahmani, M. K. I. et al., 2014).
The main techniques used in cryptography are encryption and decryption; during the encryption process, plaintext is converted into cipher text. The cipher text is converted to plain text during the decryption process. Plain text is essentially a text that contains original information that has not been encoded, whereas cipher text is the encoded text obtained immediately after the encoding of plain text (Mishra, A., 2013). A key is used to perform both data and message encryption and decryption (Chief, Z. S. et al., 2020)
Steganography involves passing confidential information through original files to keep the message hidden. The term “steganography” comes from a Greek language word that signifies “covered writing” in English. The innocent files used here may also be referred to as cover text, cover image, or cover audio, depending on the requirements (Abikoye, O. C. et al., 2012). Once a secret message implanted in a cover-image, this is referred to as stego-medium.
In this case, a stego key is normally used to control the detection and recovery processes in implanted data. Steganography is a method of concealing information within a digital image, video file, or audio file (Rashmi, N. and Jyothi, K., 2018). The basic proposed steganography-based method holds primarily two types of files. The first file is known as a cover image file, and a private key for encryption will encode the second secret file.
The encrypted secret file is then implanted in the cover image using the embedding technique, and it is compressed to minimize the number of spaces used and the data size even further. The resulting digital image, which contains a secure message hidden within it, is known as a stego image. The acquired stego image is then sent to the receiver via a computer network communication channel. Following that, using the secret key and extracting rules, the receiver will obtain the encrypted information available in the stego-image (Tyagi, M. V., 2012).
Steganography in images can be divided into two categories based on their domains: spatial domain and transform domain. Spatial domain steganography is accomplished by simply altering direct pixel values and attaching secret data to them. Some benefits, such as quick and easy implementation, make spatial domain type steganography one of the most useful steganography types (Hussain, M. et. al., 2018 and Singla, D. and Juneja, M., 2018).
Steganography techniques are the least significant among the existing spatial domain, based, palette-based, pixel value differencing, etc. However, implanting in LSB (Pramanik, S., Bandyopadhyay, S. K., 2014) is a widespread method used in spatial domain steganography (Pramanik, S. et. al., 2020).