Robust PRNU estimation from probabilistic raw measurements

作者：

Highlights：

•

• Degraded PRNU due to in-camera processing leads to poor camera identification accuracy.

• PRNU estimation from probabilistically generated raw data can improve accuracy.

• Probabilistic raw data is obtained from the JPEG output.

• Robust PRNU estimation also leads to better forgery localization.

摘要

•Degraded PRNU due to in-camera processing leads to poor camera identification accuracy.•PRNU estimation from probabilistically generated raw data can improve accuracy.•Probabilistic raw data is obtained from the JPEG output.•Robust PRNU estimation also leads to better forgery localization.

论文关键词：Multimedia forensics,Source camera identification,PhoTo Response Non Uniformity (PRNU),MLE,Tamper localization

论文评审过程：Received 7 November 2017, Revised 18 March 2018, Accepted 24 April 2018, Available online 3 May 2018, Version of Record 9 May 2018.

论文官网地址：https://doi.org/10.1016/j.image.2018.04.013