Analysis of Denoising Filters for Source Identification Using PRNU Features
In Digital Image Forensics, one of the important techniques is Source Camera Identification (SCI) that attempts to identify the source camera of captured images. The sensor patterns of captured images are used for identification. The most regular pattern noise is photo response non-uniformity (PRNU) that can be generated by sensor defects during manufacturing process. These noises are distinguishable due to different sensor vendors of devices. In this work, identification is based on mobile camera that is from which mobile model a given image is captured. Here the analysis of three different denoising filters (Wiener, Total Variation and Gaussian) are done, to get the best result in our dataset. For classification, support vector machine (SVM) classifier is used and validation is done using 10-fold cross-validation technique.
KeywordsPhoto Response non-Uniformity Wiener Total variation Gaussian
- 1.Z. Geradts, J. Bijhold, M. Kieft, K. Kurosawa, K. Kuroki, N. Saitoh, Methods for identification of images acquired with digital cameras, in Enabling Technologies for Law Enforcement and Security (2001)Google Scholar
- 10.L. Xiaolin, Based on wavelet transform plane principal component inspection application research of image denoising algorithm. Int. J. Signal Proc. Image Proc. Pattern Recogn. 8, 19–28 (2015)Google Scholar