Abstract
The issues of optimization of existing methods for digital camera sensor identification are considered. Ways of improvement of reliability of digital cameras identification is discussed. Homogeneous images were optically recorded to form a noise portrait and test sets of amateur images for 3 cameras of various types. An optimal digital filter was selected to evaluate smoothed images for obtaining noise portraits of identified cameras. Camera identification algorithm was optimized basing on a comparison of light spatial noise portraits. Application of the optimal filter and identity criterion provides an average increase of identification reliability of more than 60 times. The results can be useful in the areas of image registration and processing, security, forensics, big data analysis, etc.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
