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.

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.