Image Camera Fingerprint Extraction Algorithm Based on MPRNet

doi:10.3969/j.issn.1671-1122.2022.10.003

Abstract

Abstract:

The uniqueness and stability of Photo-Response Non-Uniformity (PRNU) noise determine can be used as a digital camera fingerprint and for traceability and forensics of digital images. In order to improve the quality of PRNU and the accuracy of image traceability, this paper proposed a PRNU camera fingerprint extraction algorithm based on multi-stage progressive image restoration network. The network uses an encoder-decoder architecture to learn contextual features and utilizes a high-resolution branch to mine local features, while introducing an attention module at each stage to reweigh local features. In this way, the global information can be fused, and the local information can be accurately obtained, so that the potential PRNU fingerprints in the natural noise of the image can be fully mined. The comparative experiments with other algorithms on the Daxing smartphone dataset and the Dresden camera dataset demonstrate the effectiveness of proposed PRNU algorithm.

Key words: image processing, image forensics, photo-response non-uniformity, source camera identification, convolutional neural network

CLC Number:

TP309

GAO Changfeng, XIAO Yanhui, TIAN Huawei. Image Camera Fingerprint Extraction Algorithm Based on MPRNet[J]. Netinfo Security, 2022, 22(10): 15-23.

Figures/Tables 11

References 22

[1]	LUKAS J, FRIDRICH J, GOLJAN M. Digital Camera Identification from Sensor Pattern Noise[J]. IEEE Transactions on Information Forensics and Security, 2006, 1(2): 205-214.
[2]	MIHCAK M K, KOZINTSEV I, RAMCHANDRAN K. Spatially Adaptive Statistical Modeling of Wavelet Image Coefficients and Its Application to Denoising[C]// IEEE. 1999 IEEE International Conference on Acoustics, Speech, and Signal Processing(ICASSP99). New York: IEEE, 1999: 3253-3256.
[3]	CORTIANA A, CONOTTER V, BOATO G, et al. Performance Comparison of Denoising Filters for Source Camera Identification[C]// SPIE. Media Watermarking, Security, and Forensics III. New York: SPIE, 2011: 60-65.
[4]	JAIN V, SEUNG H S. Natural Image Denoising with Convolutional Networks[C]// Curran Associates Inc. Advances in Neural Information Processing Systems 21. Vancouver:NeurIPS, 2008: 769-776.
[5]	BURGER H C, SCHULER C J, HARMELING S. Image Denoising: Can Plain Neural Networks Compete with BM3D[C]// IEEE. 2012 IEEE Conference on Computer Vision and Pattern Recognition. New York: IEEE, 2012: 2392-2399.
[6]	AGOSTINELLI F, ANDERSON M R, LEE H. Adaptive Multi-Column Deep Neural Networks with Application to Robust Image Denoising[C]// Curran Associates Inc. Advances in Neural Information Processing Systems 26. Vancouver:NeurIPS, 2013: 1493-1501.
[7]	XIE Junyuan, XU Linli, CHEN Enhong. Image Denoising and Inpainting with Deep Neural Networks[C]// SPIE. Advances in Neural Information Processing Systems 25. Vancouver:SPIE, 2012: 341-349.
[8]	TIAN Chunwei, XU Yong, FEI Lunke, et al. Enhanced CNN for Image Denoising[J]. CAAI Transactions on Intelligence Technology, 2019, 4(1): 17-23.
[9]	ZHANG Kai, ZUO Wangmeng, CHEN Yunjin, et al. Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising[J]. IEEE Transactions on Image Processing, 2017, 26(7): 3142-3155.
[10]	XU Jun, ZHANG Lei, ZHANG D. A Trilateral Weighted Sparse Coding Scheme for Real-World Image Denoising[C]// CVF. Proceedings of the European Conference on Computer Vision. Munich: ECCV, 2018:20-36.
[11]	ZHANG Yongsheng, TIAN Huawei, XIAO Yanhui, et al. PRNU Extraction Algorithm Based on Tripartite Weighted Sparse Coding Model[J]. Journal of Applied Science, 2019, 37(5): 704-710.
	张永胜, 田华伟, 肖延辉, 等. 基于三方加权稀疏编码模型的PRNU提取算法[J]. 应用科学学报, 2019, 37(5):704-710.
[12]	ZAMIR S W, ARORA A, KHAN S, et al. Multi-Stage Progressive Image Restoration[C]// CVF. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New York: IEEE, 2021: 14821-14831.
[13]	ZHANG Yongsheng. Research on the Technology of CNN-Based Device Source Identification and the Construction of Standard Evaluation Dataset[D]. Beijing: People’s Public Security University of China, 2020.
	张永胜. 基于CNN的图像成像设备源鉴别技术研究与标准测评库建设[D]. 北京: 中国人民公安大学, 2020.
[14]	YANG Rui, LUO Weiqi, HUANG Jiwu. Multimedia Forensics[J]. Scientia Sinica (Informationis), 2013, 43(12): 1654-1672.
	杨锐, 骆伟祺, 黄继武. 多媒体取证[J]. 中国科学:信息科学, 2013, 43(12):1654-1672.
[15]	AL-ANI M, KHELIFI F. On the SPN Estimation in Image Forensics: A Systematic Empirical Evaluation[J]. IEEE Transactions on Information Forensics and Security, 2016, 12(5): 1067-1081.
[16]	CHEN Mo, FRIDRICH J, GOLJAN M, et al. Determining Image Origin and Integrity Using Sensor Noise[C]// IEEE. 2008 IEEE Transactions on Information Forensics and Security. New York: IEEE, 2008: 74-90.
[17]	KANG Xiangui, LI Yinxiang, QU Zhenhua, et al. Enhancing Source Camera Identification Performance with a Camera Reference Phase Sensor Pattern Noise[C]// IEEE. 2012 IEEE Transactions on Information Forensics and Security. New York: IEEE, 2012: 393-402.
[18]	KANG Xiangui, LI Yinxiang, QU Zhenhua, et al. Enhancing ROC Performance of Trustworthy Camera Source Identification[C]// SPIE. Media Watermarking, Security, and Forensics III. New York: SPIE, 2011: 75-85.
[19]	LIN Xufeng, LI C T. Preprocessing Reference Sensor Pattern Noise via Spectrum Equalization[C]// IEEE. 2016 IEEE Transactions on Information Forensics and Security. New York: IEEE, 2008: 74-90.
[20]	GLOE T, BÖHME R. The Dresden Image Database for Benchmarking Digital Image Forensics[J]. Journal of Digital Forensic Practice, 2010, 3: 150-159.
[21]	TIAN Huawei, XIAO Yanhui, CAO Gang, et al. Daxing Smartphone Identification Dataset[J]. IEEE Access, 2019, 7: 101046-101053.
[22]	XIAO Yanhui, TIAN Huawei, CAO Gang, et al. Effective PRNU Extraction via Densely Connected Hierarchical Network[J]. Multimedia Tools and Applications, 2022, 81: 20443-20463.

数据集	图像大小 /像素	xDnCNN	BM3D	Wavelet	MPRNet
Daxing 数据集	128×128	0.2421	0.3660	0.4102	0.4428
	256×256	0.4826	0.5736	0.6352	0.6392
	512×512	0.6890	0.7603	0.8078	0.8022
Dresden 数据集	128×128	0.3365	0.2847	0.3411	0.4356
	256×256	0.5462	0.5241	0.6037	0.6599
	512×512	0.7373	0.7558	0.7826	0.8219