Survey of Methods used for Investigating Quantization Artefacts for Tampering Localization in Compressed Images
An image is forged when used in particular context. An image improved for fun, to remove bad photo or to improve its appearance cannot be considered as forgery even though it has been altered from its original capture. An image created to target the recipient into believing that the
image is real to achieve benefit and recognition is forgery.Here the emphasis is on passive methods for forgery detection. Blind or passive methods do not require any explicit former information about the image. A survey is made to identify the recent advancements being made in the field of digital image forgery detection. First various image forgery detection techniques are classified and then the existing blind forgery detection techniques are reviewed.This paper provides a detailed analysis of different approaches and the methodologies used to detect image forgery. It is observed that block-based features methods are robust to JPEG compression and the key point-based feature methods are robust to rotation and scaling.
Keywords -- Image Processing,Image forgery,artefacts,splicing