基于边界不确定性学习的图像篡改定位方法

doi:10.13229/j.cnki.jdxbgxb.20250014

摘要/Abstract

摘要：

针对目前图像篡改定位方法提取特征尺度单一、小篡改区域易与背景混淆造成误检漏检、预测结果不确定性高等问题，提出了基于边界不确定性学习的图像篡改定位方法。首先，使用金字塔视觉变压器提取篡改图像基础特征。其次，利用多级交互粗定位分支生成粗定位图。再次，利用小目标感知细化分支提高小篡改区域感知定位能力。随后，利用多尺度特征融合模块实现多尺度特征的充分交互与融合。最后，提出基于熵的边界不确定性感知损失进行辅助监督，极大地降低了预测结果的不确定性。在5个常用公开图像篡改数据集上分别进行域内和跨域实验，结果表明，本文方法可精准定位篡改区域，并优于其他方法。

关键词: 计算机应用, 图像篡改定位, 多尺度特征交互, 小篡改区域感知细化, 边界不确定性学习

Abstract:

The limitations of current image manipulation localization methods， such as the extraction of features at a single scale， the misdetection and omission of small tampered regions caused by background confusion， and the uncertainty existing in prediction results， are addressed. An image manipulation localization method based on edge uncertainty learning is proposed. The base features of the tampered image are extracted by means of a pyramid vision transformer. A coarse localization map is then generated through multi-level interactive coarse localization branches. To enhance the detection of small tampered regions， a small target-aware refinement branch is employed. Multi-scale feature fusion is achieved with the use of a dedicated module， which enables the full interaction and integration of features across different scales. Additionally， entropy-based perceptual loss is introduced to supervise boundary uncertainty， thus significantly reducing the uncertainty of the prediction results. The proposed method is evaluated on five widely-used public image tampering datasets in both in-domain and cross-domain experiments. It is demonstrated by the results that the method can effectively localize tampered regions and outperform existing approaches.

Key words: computer application, image manipulation localization, multi-scale feature interaction, refined perception of small tampering areas, boundary uncertainty learning

中图分类号:

TP391

陈海鹏,刘宏昕,康辉,刘雪洁. 基于边界不确定性学习的图像篡改定位方法[J]. 吉林大学学报(工学版), 2025, 55(12): 4063-4071.

Hai-peng CHEN,Hong-xin LIU,Hui KANG,Xue-jie LIU. Image manipulation localization method based on boundary uncertainty learning[J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(12): 4063-4071.

图/表 11

图1

图2

图3

图4

图5

表1

表2

表3

图6

表4

图7

参考文献 24

[1]	钟辉, 康恒, 吕颖达, 等. 基于注意力卷积神经网络的图像篡改定位算法[J]. 吉林大学学报: 工学版, 2021, 51(5): 1838-1844.
	Zhong Hui, Kang Heng, Ying-da Lyu, et al. Image manipulation localization algorithm based on channel attention convolutional neural networks[J]. Journal of Jilin University (Engineering and Technology Edition), 2021, 51(5): 1838-1844.
[2]	Shi Z, Chen H, Zhang D. Transformer-auxiliary neural networks for image manipulation localization by operator inductions[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2023, 33(9): 4907-4920.
[3]	石泽男, 陈海鹏, 张冬, 等. 预训练驱动的多模态边界感知视觉 Transformer[J]. 软件学报, 2023, 34(5): 2051-2067.
	Shi Ze-nan, Chen Hai-peng, Zhang Dong, et al. Pretraining-driven multimodal boundary-aware vision transformer[J]. Journal of Software, 2023, 34(5): 2051-2067.
[4]	Liu Y, Zhu X, Zhao X, et al. Adversarial learning for constrained image splicing detection and localization based on atrous convolution[J]. IEEE Transactions on Information Forensics and Security, 2019, 14(10): 2551-2566.
[5]	Chen B, Tan W, Coatrieux G, et al. A serial image copy-move forgery localization scheme with source/target distinguishment[J]. IEEE Transactions on Multimedia, 2020, 23: 3506-3517.
[6]	Zhang Y, Fu Z, Qi S, et al. Localization of inpainting forgery with feature enhancement network[J]. IEEE Transactions on Big Data, 2022, 9(3): 936-948.
[7]	Popescu A C, Farid H. Exposing digital forgeries in color filter array interpolated images[J]. IEEE Transactions on Signal Processing, 2005, 53(10): 3948-3959.
[8]	Bianchi T, Piva A. Image forgery localization via block-grained analysis of JPEG artifacts[J]. IEEE Transactions on Information Forensics and Security, 2012, 7(3): 1003-1017.
[9]	Mahdian B, Saic S. Using noise inconsistencies for blind image forensics[J]. Image and Vision Computing, 2009, 27(10): 1497-1503.
[10]	Chen X, Dong C, Ji J, et al. Image manipulation detection by multi-view multi-scale supervision[C]//Proceedings of the IEEE/CVF. International Conference on Computer Vision. Montreal, QC, Canada, 2021: 14165-14173.
[11]	Shi C, Wang C, Zhou X, et al. DAE-net: Dual attention mechanism and edge supervision network for image manipulation detection and localization[J]. IEEE Transactions on Instrumentation and Measurement, 2024(73): No.5028112.
[12]	Wang W, Xie E, Li X, et al. Pvt v2: Improved baselines with pyramid vision transformer[J]. Computational Visual Media, 2022, 8(3): 415-424.
[13]	Park T, Liu M Y, Wang T C, et al. Semantic image synthesis with spatially-adaptive normalization[C]∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach, CA: USA, 2019: 2337-2346.
[14]	Pang Y, Zhao X, Xiang T Z, et al. Zoom in and out: A mixed-scale triplet network for camouflaged object detection[C]∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans, LA: USA, 2022: 2160-2170.
[15]	Dong J, Wang W, Tan T. Casia image tampering detection evaluation database[C]∥2013 IEEE. China Summit and International Conference on Signal and Information Processing.Beijing: China, 2013: 422-426.
[16]	Guan H, Kozak M, Robertson E, et al. MFC datasets: Large-scale benchmark datasets for media for ensic challenge evaluation[C]∥2019 IEEE.Winter Applications of Computer Vision Workshops (WACVW). Waikoloa, HI: USA, 2019: 63-72.
[17]	Wen B, Zhu Y, Subramanian R, et al. COVERAGE—a novel database for copy-move forgery detection[C]∥2016 IEEE. International Conference on Image Processing (ICIP).Phoenix, AZ: USA, 2016: 161-165.
[18]	Hsu Y F, Chang S F. Detecting image splicing usinggeometry invariants and camera characteristics consistency[C]∥2006 IEEE. International Conference on Multimedia and Expo. Toronto, ON: Canada, 2006: 549-552.
[19]	Novozamsky A, Mahdian B, Saic S. IMD2020 a largescale annotated dataset tailored for detecting manipulated images[C]∥Proceedings of the IEEE/CVF. Winter Conference on Applications of Computer Vision Workshops. Snowmass Village, CO: USA, 2020: 71-80.
[20]	Wu Y, AbdAlmageed W, Natarajan P. Mantra-net manipulation tracing network for detection and localization of image forgeries with anomalous features[C]∥Proceedings of the IEEE/CVF. Conference on Computer Vision and Pattern Recognition. Long Beach, CA: USA, 2019: 9543-9552.
[21]	Hu X, Zhang Z, Jiang Z, et al. SPAN spatial pyramid attention network for image manipulation localization[C]∥Proceedings of the European Conference on Computer Vision (ECCV). Glasgow: UK, 2020: 312-328.
[22]	Zhuang P, Li H, Tan S, et al. Image tampering localization using a dense fully convolutional network[J]. IEEE Transactions on Information Forensics and Security, 2021, 16: 2986-2999.
[23]	Zhuo L, Tan S, Li B, et al. Self-adversarial training incorporating forgery attention for image forgery localization[J]. IEEE Transactions on Information Forensics and Security, 2022, 17: 819-834.
[24]	Xia X, Su L C, Wang S P, et al. DMFF-net: Double-stream multilevel feature fusion network for image forgery localization[J]. Engineering Applications of Artificial Intelligence, 2024, 127: No.107200.

相关文章 15

[1]	艾青林,刘元宵,杨佳豪. 基于MFF-STDC网络的室外复杂环境小目标语义分割方法[J]. 吉林大学学报(工学版), 2025, 55(8): 2681-2692.
[2]	张宇飞,王丽敏,赵建平,贾智尧,李明洋. 基于中心选择大逃杀优化算法的机器人逆运动学求解[J]. 吉林大学学报(工学版), 2025, 55(8): 2703-2710.
[3]	朴燕,康继元. RAUGAN：基于循环生成对抗网络的红外图像彩色化方法[J]. 吉林大学学报(工学版), 2025, 55(8): 2722-2731.
[4]	刘琼昕,王甜甜,王亚男. 非支配排序粒子群遗传算法解决车辆位置路由问题[J]. 吉林大学学报(工学版), 2025, 55(7): 2464-2474.
[5]	车翔玖,李良. 融合全局与局部细粒度特征的图相似度度量算法[J]. 吉林大学学报(工学版), 2025, 55(7): 2365-2371.
[6]	李文辉,杨晨. 基于对比学习文本感知的小样本遥感图像分类[J]. 吉林大学学报(工学版), 2025, 55(7): 2393-2401.
[7]	庄珊娜,王君帅,白晶,杜京瑾,王正友. 基于三维卷积与自注意力机制的视频行人重识别[J]. 吉林大学学报(工学版), 2025, 55(7): 2409-2417.
[8]	赵宏伟,周伟民. 基于数据增强的半监督单目深度估计框架[J]. 吉林大学学报(工学版), 2025, 55(6): 2082-2088.
[9]	王健,贾晨威. 面向智能网联车辆的轨迹预测模型[J]. 吉林大学学报(工学版), 2025, 55(6): 1963-1972.
[10]	陈海鹏,张世博,吕颖达. 多尺度感知与边界引导的图像篡改检测方法[J]. 吉林大学学报(工学版), 2025, 55(6): 2114-2121.
[11]	周丰丰,郭喆,范雨思. 面向不平衡多组学癌症数据的特征表征算法[J]. 吉林大学学报(工学版), 2025, 55(6): 2089-2096.
[12]	车翔玖,孙雨鹏. 基于相似度随机游走聚合的图节点分类算法[J]. 吉林大学学报(工学版), 2025, 55(6): 2069-2075.
[13]	刘萍萍,商文理,解小宇,杨晓康. 基于细粒度分析的不均衡图像分类算法[J]. 吉林大学学报(工学版), 2025, 55(6): 2122-2130.
[14]	王友卫,刘奥,凤丽洲. 基于知识蒸馏和评论时间的文本情感分类新方法[J]. 吉林大学学报(工学版), 2025, 55(5): 1664-1674.
[15]	赵宏伟,周明珠,刘萍萍,周求湛. 基于置信学习和协同训练的医学图像分割方法[J]. 吉林大学学报(工学版), 2025, 55(5): 1675-1681.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

数据集	分割设置		篡改操作类型			后处理操作
数据集	训练	测试	拼接	复制-粘贴	移除	后处理操作
CASIA^［15］	5 123	921	√	√		√
NIST16^［16］	404	160	√	√	√	√
COVER^［17］	75	25		√		√
Columbia^［18］	130	50	√
IMD2020^［19］	1 610	400	√	√	√	√

方法	CASIAV		NIST16		Columbia		COVER		IMD2020		Mean
方法	AUC	F₁	AUC	F₁	AUC	F₁	AUC	F₁	AUC	F₁	AUC	F₁
ManTra-Net^［20］	0.648	0.223	0.795	0.462	0.824	—	0.777	0.283	0.785	0.265	0.766	0.308
SPAN^［21］	0.709	0.213	0.961	0.582	0.936	0.815	0.791	0.325	—	—	0.849	0.484
DenseFCN^［22］	0.631	0.209	0.954	0.704	0.881	0.710	0.754	0.185	0.723	0.286	0.789	0.419
SATFL^［23］	0.697	0.246	0.937	0.613	0.892	0.804	0.767	0.347	0.796	0.300	0.818	0.462
MVSS-Net^［10］	0.748	0.390	0.981	0.827	0.719	0.703	0.808	0.284	0.817	0.411	0.815	0.523
TANet^［2］	0.739	0.493	0.955	0.901	0.982	0.960	0.756	0.425	0.766	0.383	0.840	0.633
DMFF-Net^［24］	0.791	0.386	0.976	0.843	0.945	0.837	0.727	0.282	0.784	0.328	0.845	0.535
DAE-Net^［11］	0.805	0.494	0.984	0.871	0.973	0.872	0.741	0.330	0.826	0.338	0.866	0.581
PBUL-Net	0.852	0.662	0.981	0.906	0.982	0.961	0.819	0.525	0.796	0.488	0.886	0.708

方法	CASIAV		NIST16		Columbia
方法	AUC	F₁	AUC	F₁	AUC	F₁
ManTra-Net^［21］	0.648	0.223	0.475	0.095	0.481	0.386
DenseFCN^［22］	0.631	0.209	0.604	0.051	0.586	0.317
MVSS-Net^［23］	0.748	0.390	0.643	0.246	0.697	0.471
TANet^［2］	0.739	0.493	0.638	0.248	0.710	0.440
DAE-Net^［12］	0.805	0.494	0.663	0.299	0.726	0.486
PBUL-Net	0.852	0.662	0.702	0.344	0.800	0.690

网络	组件					AUC	F₁
网络	MFFM	X	FAM	FAEM	EUBAL	AUC	F₁
a						0.810	0.605
b	√					0.819	0.618
c	√	√				0.832	0.629
d	√	√	√			0.838	0.641
e	√	√	√	√		0.848	0.653
f	√	√	√	√	√	0.852	0.662