LSB Steganography in Python: Hide Data in Images 🖼️

LSB Steganography - Hiding secret data (a low-resolution Zelda image) at different bitplanes inside a cover image (Lena) with different Number Systems (Binary, Prime and Natural Number Systems) and extracting the secret image from the stego image back. Prime / Natural Number Systems are defined here to be binary representations of pixel values with prime / natural numbers as basis, respectively.

Since the number systems with Prime / Natural numbers as basis don't have unique representation (the mapping is not injective / non-invertible) of pixel values, not all pixel values are embeddable (only the red pixels constructing the embeddable mask contain the secret data bits embedded, the rest of the pixels from cover image are ignored and not used for embedding).

As it can be seen, with Prime and Natural Number Systems more secret data can be hidden to higher bitplanes inside the input cover image with less visible distortion in the output stego image (as opposed to standard binary representation with only 8 bitplanes available, the prime and natural number systems have 15 and 24 bitplanes, respectively, for an 8-bit gray scale cover image).

More here: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.300.3753 or here: https://www.tifr.res.in/~sanyal/papers/Sandipan_Datahiding_PrimeNatural.pdf

#imageprocessing #imageprocessingpython #steganography #python #numbersystem #primenumbers