Coarse-to-fine-grained method for image splicing region detection

作者：

Highlights：

• Interpreting the difference of CFA interpolation pattern as an evidence of splicing.

• Apply the forensics features to obtain the coarse-grained detection results.

• Use local texture strength features to perform fine-grained detection.

• Use a method named edge smoothing to realize precise localization.

摘要

•Interpreting the difference of CFA interpolation pattern as an evidence of splicing.•Apply the forensics features to obtain the coarse-grained detection results.•Use local texture strength features to perform fine-grained detection.•Use a method named edge smoothing to realize precise localization.

论文关键词：Image splicing detection,CFA interpolation algorithm,Forensics features,Texture strength features,Edges smoothing

论文评审过程：Received 18 February 2021, Revised 16 September 2021, Accepted 20 September 2021, Available online 22 September 2021, Version of Record 1 October 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2021.108347