For grayscale images use 'gray_encryption.py' for hiding and 'gray_decryption.py' for decoding. The cover image name must be named "Cover_1.png".
For RGB images use 'color_encryption.py' for hiding and 'color_decryption.py' for decoding. The cover image name must be named "Cover_2.png".
The algorithm is taken from the paper 'A Secure Robust Gray Scale Image Steganography Using Image Segmentation'
The algorithm is as follows:
Now, let us expand how the Steganography algorithm works for ith bit:
- l = number of rows, r = number of columns, klen = key length
- Let x = 2nd LSB of C(i, j), where C(i, j) is the element on ith row and jth column of the cover image.
- Let mbit = M(cnt) {i.e. cntth bit of message}. Let k = K(i, j), i.e. the respective bit from key K. k can be calculated as: k = key[(i*r + j)%klen]
- if(x^k = 1) then a) Change LSB of that image element b) Increase cnt by 1 c) Check the 2nd and 3rd LSB of the Cover image pixel. Let g = string(2nd LSB) + string(3rd LSB), where ‘+’ is concatenation operator. Possible g values are {00, 01, 10, 11} d) If the LSB of the pixel was changed in the step 4a, then increase the change-count of that particular g-type. Else increase the not-changed-count by 1. For example if g=00 and LSB was changed from 0 to 1, then nc00 = nc00+1. If LSB wasn’t changed, c00 = c00+1
- Move to the inversion algorithm part.
Let us move on to the inversion algorithm:
- For each pixel S(i, j) of the stego-image: a) Check the 2nd and 3rd LSB of the Cover image pixel. Let g = string(2nd LSB) + string(3rd LSB), where ‘+’ is concatenation operator. Possible g values are {00, 01, 10, 11} b) If the changed-count>not-changed-count for that particular g, invert the LSB of that pixel. For example: if g = 00, nc00 = 200 and c00=250, LSB = 0 then invert the LSB from 0 to 1.
- Compute p. p is embedded in the LSBs of the first 4 pixels. It signifies whether inversion is done or not for the four categories of g: {00, 01, 10, 11}. p[0] = (nc00>c00) p[1] = (nc01>c01) p[2] = (nc10>c10) p[3] = (nc11>c11)
The DECRYPTION ALGORITHM is as follows:
- Concatenate the LSBs of the first 4 pixels to form p.
- Initialize msg = “”
- Then loop over all pixels except the first 4 : a. Let x = 2nd LSB of S(i, j), where C(i, j) is the element on ith row and jth column of the cover image. b. Let k = K(i, j), i.e. the respective bit from key K. k can be calculated as: k = key[(i*r + j)%klen] c. if(x^k = 1) then: i. msg = msg + getData(S(I, j), p). ii. Here, the getData(pixel_val,p) function extracts the appropriate bit from the pixel. We compute g as we did in the encryption algorithm, then either invert the LSB(if the respective bit in p=1) or return the LSB as it is(otherwise).
- We then extract the first 17 bits of msg, convert it from binary to decimal, and get the size of the secret message. mlen = msg[0:17] mlen = int(mlen)
- Now we extract the required bits from msg (size = mlen) msg = msg[17:17+mlen]
- Finally, we convert the message from binary to ASCII
The algorithm is taken from the paper 'A Steganography Algorithm for Hiding Secret Message inside Image using Random Key'.
Encryption Algorithm:
Decryption Algorithm:
