Journal of Computer Applications ›› 2023, Vol. 43 ›› Issue (9): 2919-2924.DOI: 10.11772/j.issn.1001-9081.2022081288

• Multimedia computing and computer simulation • Previous Articles     Next Articles

Image copy-move forgery detection based on multi-scale feature extraction and fusion

Juntao CHEN, Ziqi ZHU()   

  1. School of Computer Science and Technology,Wuhan University of Science and Technology,Wuhan Hubei 430081,China
  • Received:2022-08-30 Revised:2022-11-09 Accepted:2022-11-14 Online:2023-01-11 Published:2023-09-10
  • Contact: Ziqi ZHU
  • About author:CHEN Juntao, born in 1992, M. S. candidate. His research interests include computer vision, forgery detection.
  • Supported by:
    National Natural Science Foundation of China(61702382)


陈俊韬, 朱子奇()   

  1. 武汉科技大学 计算机科学与技术学院,武汉 430081
  • 通讯作者: 朱子奇
  • 作者简介:陈俊韬(1992—),男,福建福州人,硕士研究生,主要研究方向:计算机视觉、伪造检测;
  • 基金资助:


In the field of image copy-move forgery detection, it is very challenging to locate the boundaries of tampered small objects accurately. Current deep learning-based methods locate forged regions by detecting similar content in images. However, these methods usually just transmit the final features extracted by the encoder to the decoder to generate the mask, and ignore more spatial information of forged regions contained in the high-resolution encoding features, resulting in inaccurate model output prediction results for the boundary identification of small objects. To address this problem, a detection network based on multi-scale feature extraction and fusion called SimiNet was proposed. Firstly, abundant features were extracted by the multi-scale feature extraction module. Secondly, a skip connection was added between the feature extraction module and the decoding module to bridge the gap between the encoding and decoding features, so as to identify the boundaries of small objects accurately. Finally, Log-Cosh Dice Loss function was used to take the place of cross entropy loss to reduce the impact of class-imbalance problem on detection results. Experimental results show that the F1 score of SimiNet on USCISI dataset reaches 72.54%, which is 3.39 percentage points higher than that of the suboptimal method CMSDNet (Copy-Move Similarity Detection Network). It can be seen that SimiNet is more accurate for boundary identification of small objects and has better visualization.

Key words: copy-move, forgery detection, deep learning, skip connection, similarity detection


在图像复制-粘贴伪造检测领域,精确地定位被篡改的小目标的边界充满挑战性。当前基于深度学习的方法通过检测图像中的相似内容来定位伪造区域,然而它们通常只是把编码器最终提取的特征传递给解码器来生成掩码,忽略了高分辨率的编码特征所包含的更多的伪造区域的空间信息,这会导致模型输出的预测结果对于小目标的边界识别不精确。针对这个问题,提出一种基于多尺度特征提取与融合的检测网络SimiNet。首先,使用多尺度特征提取模块提取丰富的特征;其次,在特征提取模块与解码模块之间添加跳跃连接以弥补编码特征与解码特征之间的差异,从而精确地识别小目标的边界;最后,用Log-Cosh Dice Loss函数替代交叉熵损失,以降低类别不平衡问题对检测结果的影响。实验结果表明,SimiNet在USCISI数据集上的F1分数达到72.54%,比次优方法CMSDNet(Copy-Move Similarity Detection Network)提升了3.39个百分点。可见,SimiNet对小目标的边界识别更精确,可视化效果更好。

关键词: 复制粘贴, 伪造检测, 深度学习, 跳跃连接, 相似性检测

CLC Number: