Detection and Localization Of Digital Image Forgeries (eBook, PDF)
Versandkostenfrei!
Sofort per Download lieferbar
39,99 €
inkl. MwSt.
PAYBACK Punkte
0 °P sammeln!
Doctoral Thesis / Dissertation from the year 2018 in the subject Engineering - Computer Engineering, grade: 8.83, , language: English, abstract: The use of digital images has increased rapidly over the past few years and so have the means and the incentive to create digital forgeries. Presently, powerful editing software allows forgers to create perceptually convincing forgeries. The manipulation of images through forgery influences the perception an observer has of the depicted scene, potentially resulting in ill consequences if created with malicious intentions. This poses a need to verify t...
Doctoral Thesis / Dissertation from the year 2018 in the subject Engineering - Computer Engineering, grade: 8.83, , language: English, abstract: The use of digital images has increased rapidly over the past few years and so have the means and the incentive to create digital forgeries. Presently, powerful editing software allows forgers to create perceptually convincing forgeries. The manipulation of images through forgery influences the perception an observer has of the depicted scene, potentially resulting in ill consequences if created with malicious intentions. This poses a need to verify the authenticity of images before it is put to use for making any critical decision. The work in the area of image content authentication has widely focused on active authentication methods like digital signatures and digital watermarks. These methods, however suffer from the drawback that they need to be inserted at the time of recording the image. Insertion of watermarks and signatures into the image require special hardware like specially equipped cameras and as such limit the use of these approaches. To overcome this drawback passive authentication has emerged as an alternative. Passive authentication needs no information about the image just the image itself. The passive domain is classified into image source identification and tamper detection. Image source identification takes advantage of the traces left by the different processing steps in the image acquisition and storage phases. These traces mark the image with some kind of camera fingerprint, which can be used for authentication. However to forge the source of an image a forger need to add noise of a specific camera to the image, remove its pattern noise or alter chromatic abrasions. All this requires a forger to master the image acquisition process as well as the editing process of the source information and thus makes it is less suitable in a wider domain. Thus tamper detection is a tool applicable in a larger set of practical situations since tampering with image requires no expert knowledge. Considering the importance of image authentication and the ease with which an image can be manipulated this thesis attempts to detect if an image has been tampered, to check its authenticity, in case the image has been tampered a step further an attempt has been made to determine location of copy-move forgery and extent of doctoring.
Andere Kunden interessierten sich für
