Skip to content

Latest commit

 

History

History
15 lines (9 loc) · 786 Bytes

File metadata and controls

15 lines (9 loc) · 786 Bytes

Digital camera identification tool

The implementation is based on Digital Camera Identification from Sensor Pattern Noise (Lukáš, Fridrich and Goljan). My multimedia forensics bibliography lists more papers on this topic.

Prerequisites:

  • pip install pillow cython pywavelets
  • Run python setup.py build_ext in accel, then link the .so output into that directory.

make_characteristic.py

Given a path containing PNG-format images, calculates the characteristic noise signal and outputs it to a file.

test_characteristic.py

Outputs the correlation of a given characteristic noise signal and test image(s).