What are the methods of Image Steganography in Information Security?

Image steganography has truly state-of-the-art with the invention of quick, dynamic computers. Software is simply accessible for processing and hiding of data images. Images can also be recovered very simply.

There are three main methods of information hiding in images which are as follows −

Least Significant Bit Insertion − This is the most famous image steganography technique. It is simple, easy to make and also simply to use. Unfortunately, it is intensely vulnerable to attack. A simple conversion of structure can damage all hidden data.

The best approach of image document to hide data inside is a 24 Bit BMP (Bitmap) image. When an image is of large quality and resolution it is simpler to hide data within image. Although 24 Bit images are best for hiding data because of their size.

Some people can select 8 Bit BMP‟s or possibly another image format including GIF. The important reason being is that posting of high images on the internet can stimulate suspicion.

The least significant bit i.e. the eighth bit can be used to change to a bit of the private message. When using a 24-bit image, one can save 3 bits in every pixel by transforming a bit of each of the red, green and blue color element.

Masking and filtering − Masking and filtering hide data by marking an image in a manner same as paper watermarks. By masking a faint image with another in order to create the first non-perceptible, and it can exploit the fact that the human eye cannot identify faint changes in a visual picture.

Masking approach are more acceptable for use in lossy JPEG images than in LSB insertion because of their associative immunity to condensation and cropping.

Algorithms and transformation − JPEG images are very popular, as they provide high quality pictures and the ability to hide information in them. JPEG uses the discrete cosine transform (DCT) to achieve compression.

DCT is a lossy transform, as cosine values cannot be computed exactly and rounding errors frequently take place. Variances in data based on the approach and values used.

Images can also be handled using Fourier transformations and wavelet transformations. Spread spectrum can be used, as a narrow bandwidth can be spread over a higher one in such a method that the signal’s spectral density looks like noise.

Direct-sequence and frequency-hopping spread-spectrum approaches are used. Direct-sequencing works by phase-modulating the information with a pseudorandom number sequence that both the sender and receiver are receptive of.

Frequency-hopping compromise with dividing the bandwidth into several channels and then hopping among them. There are other techniques encrypt and discard the message throughout the image in some pre-decided manner.

It is considered that even if the message is disclosed, it is meaningless without the algorithm and the key. Unfortunately, none of these algorithms are invulnerable to destruction of information through image manipulation.

Updated on: 11-Mar-2022


Kickstart Your Career

Get certified by completing the course

Get Started