prnu

Python implementation of PRNU extractor

Implementation of the famous Camera Noise Fingerprint "NoisePrint" in Pytorch

A Python script that extracts the PRNU from images using the DRUNet network as a denoising technique.

Web app for image forgery detection. Runs 12 published forensic techniques: ELA, Splicebuster noise analysis, double-JPEG and JPEG ghosts, copy-move, PRNU camera fingerprint, resampling, histogram, LSB steganalysis, metadata/C2PA and more. Reports what each test measured and its limits, never a black-box real-or-fake verdict. Includes demo images.

This repository provides a PyTorch implementation of FFDNet image denoising https://arxiv.org/abs/1710.04026. First implemented by Matias Tassano https://doi.org/10.5201/ipol.2019.231, the FFDNet source code has been adapted so as to extract cameras PRNU.

PRNU extraction + validation

超级耿同学 — 学术论文多维度异常检测框架。不简单给二值判断，用贝叶斯证据网络对每条证据建模，输出带不确定性的后验造假概率。理工科7维+文科5维+因果推断+图像法医。

2016JFS, Source camera identification with content adaptive guided filter

IEEE's Signal Processing Society - Camera Model Identification Identify from which camera an image was taken

Python implementation of PRNU extractor

Zero-knowledge circuits for sensor-based photo provenance (Noir / UltraHonk)

The Video Truth Record (.vtr) is an open standard for hardware-attested media. It cryptographically binds video files to the unique physical signature of the camera sensor (PRNU) at capture, utilizing zero-knowledge proofs to guarantee authenticity and an unbroken chain of custody.

Wavelet-domain PRNU based image authentication system for verifying whether an image originates from an enrolled camera sensor.

Deep Learning framework for Source Camera Identification using PRNU (Sensor Pattern Noise), Wavelet Residual Extraction, CNN Feature Learning, and Hybrid Feature Fusion.

Offline photo-forensics tool: metadata, stripped-photo clues, ELA, copy-move, PRNU camera fingerprinting. No AI, no API keys, no network.

AI-media forensics platform: multi-tier detection, edit localization, claim-bundle sensor fingerprinting, calibrated court-defensible error rates.

Python program for analyzing the device using which the picture was taken.