An improved parameter estimation scheme for image modification detection based on DCT coefficient analysis
详细信息 查看全文
- 作者：Liyang Yua ; b ; Qi Hana ; qi.han@hit.edu.cn" class="auth_mail" title="E-mail the corresponding author ; Xiamu Niua ; xm.niu@hit.edu.cn" class="auth_mail" title="E-mail the corresponding author ; S.M. Yiuc ; Junbin Fangd ; Ye Zhange
- 关键词：Digital image forensics ; Image forgery detection ; DCT coefficient analysis ; Primary quantization step estimation
- 刊名：Forensic Science International
- 出版年：2016
- 出版时间：February 2016
- 年：2016
- 卷：259
- 期：Complete
- 页码：200-209
- 全文大小：2695 K
文摘
- •
A coarse-to-fine parameter estimation scheme for image modification detection.
- •
The smoothness property of the model motivates the coarse-grained search.
- •
The optimal value is approached by gradient ascent.
- •
The constraint for the parameter determines the learning rate.