conseal

conseal is a Python package, containing implementations of modern image steganographic algorithms.

Note

This project is under active development.

Simulating steganography has never been easier!

>>> # load JPEG cover
>>> import jpeglib
>>> im_dct = jpeglib.read_dct("cover.jpeg")
>>> im_px = jpeglib.read_spatial("cover.jpeg")
>>>
>>> # simulate JPEG steganography
>>> import conseal as cl
>>> im_dct.Y = cl.juniward.simulate_single_channel(
...   cover_dct_coeffs=im_dct.Y,
...   quantization_table=im_dct.qt[0],
...   cover_spatial=im_px.spatial[..., 0],
...   embedding_rate=0.4,
...   seed=12345)
...
>>> # save stego
>>> im_dct.write_dct("stego.jpeg")

>>> # load spatial cover
>>> import numpy as np
>>> from PIL import Image
>>> cover_spatial = np.array(Image.open("cover.png"))
>>>
>>> # simulate spatial steganography
>>> import conseal as cl
>>> stego_spatial = cl.lsb.simulate(
...   cover=cover_spatial,
...   embedding_rate=0.4,
...   seed=12345)
...
>>> # save stego
>>> Image.fromarray(stego_spatial).save("stego.png")
Available embedding simulators.

Domain

Algorithms

DCT

J-UNIWARD, UERD, EBS, nsF5, LSB

Spatial

LSB

Contents