Detection of Image Region Duplication Using Spin Image

スポンサーリンク

概要

• 論文の詳細を見る

• Passive-blind image forensics is a technique that judges whether an image is forged in the absence of watermarking. In image forgery, region duplication is a simple and widely used method. In this paper, we proposed a novel method to detect image region duplication using the spin imagewhich is an intensity-based and rotation invariant descriptor. The method can detect region duplication exactly and is robust to geometric transformations. Furthermore, it is superior to the popular SIFT-based detection method when the copied patch is from smooth background. The experiments have proved the method's effectiveness.

著者

• Niu Xiamu

  Department Of Computer Science And Technology Harbin Institute Of Technology

• JIANG Lin

  Harbin Institute of Technology Shenzhen Graduate School

• WANG Shen

  Department of Computer Science and Technology, Harbin Institute of Technology

• SONG Xianhua

  Department of Computer Science and Technology, Harbin Institute of Technology

• YIU Siuming

  Department of Computer Science, University of Hong Kong

関連論文

• MHD vortices in the magnetospheric low latitude boundary layer
• A Novel Bayes' Theorem-Based Saliency Detection Model
• Drain/Gate-Voltage-Dependent On-Current and Off-Current Instabilities in Polycrystalline Silicon Thin-Film Transistors under Electrical Stress
• Detection of Image Region Duplication Using Spin Image
• Region Diversity Based Saliency Density Maximization for Salient Object Detection
• CBRISK: Colored Binary Robust Invariant Scalable Keypoints
• Saliency Density and Edge Response Based Salient Object Detection

スポンサーリンク