Improving sensor noise analysis for CT-Scanner identification

Anas Kharboutly 1 William Puech 1 Gérard Subsol 1 Denis Hoa 2
1 ICAR - Image & Interaction
LIRMM - Laboratoire d'Informatique de Robotique et de Microélectronique de Montpellier
2 IMAIOS
Abstract : CT-Scanner devices produce three-dimensional images of the internal structure of the body. In this paper, we propose a method that is based on the analysis of sensor noise to identify the CT-Scanner device. For each CT-scanner we built a reference pattern noise and a correlation map from its slices. Finally, we can correlate any test slice with the reference pattern noise of each device according to its correlation map. This correlation map gives a weighting for each pixel regarding its position in the reference pattern noise. We used a wavelet-based Wiener filter and an edge detection method to extract the noise from a slice. Experiments were applied on three CT-Scanners with 40 3D images, including 3600 slices, and we demonstrate that we are able to identify each CT-Scanner separately.
Keywords : Digital forensics Medical image forensics Authentication Device identification Noise pattern Sensor noise Denoise filtering Edge detection Wavelet transformation
Document type :
Conference papers
Complete list of metadatas

Cited literature [18 references]  Display  Hide  Download
https://hal-lirmm.ccsd.cnrs.fr/lirmm-01379558
Contributor : William Puech <>
Submitted on : Wednesday, March 27, 2019 - 2:16:23 PM
Last modification on : Wednesday, March 27, 2019 - 2:28:22 PM
Long-term archiving on: Friday, June 28, 2019 - 2:34:45 PM

File

07362817.pdf
Publisher files allowed on an open archive

Identifiers

Collections

LIRMM | CNRS | ICARLIRMM | MIPS | UNIV-MONTPELLIER

Citation

Anas Kharboutly, William Puech, Gérard Subsol, Denis Hoa. Improving sensor noise analysis for CT-Scanner identification. EUSIPCO: European Signal Processing Conference, Aug 2015, Nice, France. pp.2411-2415, ⟨10.1109/EUSIPCO.2015.7362817⟩. ⟨lirmm-01379558⟩

Export

BibTeX TEI DC DCterms EndNote

Share

Metrics

Record views

129

Files downloads

54

Contact

Informations

CCSD