发明名称 |
Building a digital camera fingerprint from cropped or corrupted images |
摘要 |
A method of identifying a digital camera is disclosed. First and second digital images generated by the camera have dimensions that are not equal. The method includes calculating noise residuals and normalized cross-correlations (NCCs) between the noise residuals corresponding to candidate translations. The method further includes calculating a noise floor and identifying first and second peak values corresponding to translations. The method further includes calculating a peak ratio, determining that the peak ratio exceeds an alignment threshold and calculating a digital camera fingerprint for the digital camera based on the noise residuals and the first translation. The method further includes receiving a digital image, calculating a noise residual of the digital image, and determining that the image was generated by the digital camera based on the noise residual and the fingerprint and generating a record associating the image with the camera. |
申请公布号 |
US9525866(B1) |
申请公布日期 |
2016.12.20 |
申请号 |
US201615131192 |
申请日期 |
2016.04.18 |
申请人 |
The United States of America as represented by the Director, National Security Agency |
发明人 |
Charlton Sarah T.;Martin Carla D. |
分类号 |
G06K9/00;H04N17/00;G06T5/00 |
主分类号 |
G06K9/00 |
代理机构 |
|
代理人 |
|
主权项 |
1. A computer-implemented method of using a digital camera fingerprint to identify a digital camera that generated a test digital image, the method comprising:
receiving a first digital image generated by the digital camera and having first dimensions; receiving a second digital image generated by the digital camera and having second dimensions, wherein the first dimensions and the second dimensions are not equal; calculating a first noise residual of the first image; calculating a second noise residual of the second image; calculating a plurality of normalized cross-correlations (NCCs) between the first noise residual and the second noise residual corresponding to a plurality of candidate translations; calculating a noise floor based on the plurality of NCCs; identifying a first peak value based on the plurality of NCCs and the noise floor and corresponding to a first translation; identifying a second peak value based on the plurality of NCCs and the noise floor and corresponding to a second translation; calculating a peak ratio based on the first peak value and the second peak value; determining that the peak ratio exceeds an alignment threshold; calculating a digital camera fingerprint for the digital camera based on the first noise residual, the second noise residual, and the first translation; receiving a third digital image generated by the digital camera; calculating a third noise residual of the third image; determining a third translation having an associated peak ratio exceeding the alignment threshold; calculating an updated digital camera fingerprint for the digital camera based on the digital camera fingerprint, the third noise residual, and the third translation; receiving the test digital image; calculating a test noise residual of the test digital image; determining that the test digital image was generated by the digital camera based on the test noise residual and the updated digital camera fingerprint; and generating a digital record associating the test digital image with the digital camera. |
地址 |
Washington DC US |