DEFENDING MECHANISM AGAINST FINGER PRINT ATTACK USING DYNAMIC TIME WARPING ALGORITHM
Authors: Dr. R. Priyadarshini, Dr. K. Aravindhan, Dr. M. Sowmya
DOI: 10.87349/JBUPT/27704
Page No: 26-34
Abstract
Photo-response non-uniformity (PRNU) of imaging sensors can be employed as a distinctive fingerprint to deal with diverse forensic tasks connecting digital images and video. One of the most significant applications of this knowledge is matching an image or a video snip to the camera that acquired it, which is a task analogous in spirit to matching a shell to a gun cask. The problem examined here distress the condition when an adversary guesstimates the sensor fingerprint from a set of images and applies to it onto an image from a diverse camera to enclose an innocent victim. The previous paper provides a consistent method for detecting such fake fingerprints under quite mild and common assumptions regarding the adversary’s movement and the means accessible to the victim. The major drawback of the previous work is that every attack is guessed based on defenders’ assumptions. To overcome this, in this paper, we plan to present Dynamic Time Warping (DTW) algorithm to detect sensor finger print. DTW identify similarity among two sequences of finger print images alter in time or speed can be applied to video, audio, and graphics with linear representation. Evaluate an optimal match between two given sequences to detect fake sensor fingerprint at the specific time and event. Sequences are warped non-linearly at the time of fake infusion Utilized with hidden Markov models. Fake Sensor fingerprint detection time is identified exactly by quantifying the similarity variance. Source of Fake generated camera features can be approximately identified. An experimental evaluation is conducted with set of tests in real system to estimate the performance of the proposed DTW to detect sensor finger print contarst with an existing Defending Against Fingerprint-Copy Attack in Sensor-Based Camera Identification.
