Image splicing forgery detection using simplified generalized noise model

作者：

Highlights：

• A simplified generalized noise model characterizing a natural JPEG image is proposed.

• A model-based forensic detector is devised to detect image splicing forgery.

• The statistical performance of detector can be analytically established.

• An effective fusion strategy is proposed to improve the performance of detector.

摘要

•A simplified generalized noise model characterizing a natural JPEG image is proposed.•A model-based forensic detector is devised to detect image splicing forgery.•The statistical performance of detector can be analytically established.•An effective fusion strategy is proposed to improve the performance of detector.

论文关键词：Image splicing detection,Noise model,Camera fingerprints,Hypothesis testing

论文评审过程：Received 20 May 2020, Revised 16 May 2022, Accepted 13 June 2022, Available online 18 June 2022, Version of Record 1 July 2022.

论文官网地址：https://doi.org/10.1016/j.image.2022.116785