NUST Institutional Repository

CONTENT BASED FRAGILE WATERMARKING SCHEME FOR IMAGE AUTHENTICATION AND ERROR LOCALIZATION

Show simple item record

dc.contributor.author RAHMAN, MUFTI TAUSIF-UR-
dc.date.accessioned 2023-08-29T06:08:50Z
dc.date.available 2023-08-29T06:08:50Z
dc.date.issued 2009
dc.identifier.other 2005-NUST-MSc PhD- ComE-05
dc.identifier.uri http://10.250.8.41:8080/xmlui/handle/123456789/37786
dc.description Supervisor: DR MUHAMMAD YOUNUS JAVED en_US
dc.description.abstract With the availability of some powerful image processing softwares such as ‘Adobe Photoshop’, one can remove/replace some features in a picture easily without any detectable trace. This type of operation may be regarded as tamper. However, in some cases, a person or an organization cannot afford an image to undergo any such operation, such as images for military, medical or judicative use etc. The validity and authenticity of an image is of utmost importance in such cases, so there is a need to guarantee the integrity of an image in an effective manner. Also in the second phase, one must be able to localize and identify an area which has been tampered or has developed some sort of an error while passing the transmission channel or during storage/retrieval process. This thesis presents an image authentication and error localization technique, based on fragile water marking. The watermark is content based and the error localization method uses a “Cyclic Redundancy Checksum” to authenticate image as well as localize any channel induced or intentionally introduced error, down to block level (2x2 pixels). The scheme has been tested on various image formats as well as image sizes and has successfully localized error/tampering with good PSNR valves. en_US
dc.language.iso en en_US
dc.publisher College of Electrical & Mechanical Engineering (CEME), NUST en_US
dc.title CONTENT BASED FRAGILE WATERMARKING SCHEME FOR IMAGE AUTHENTICATION AND ERROR LOCALIZATION en_US
dc.type Thesis en_US


Files in this item

This item appears in the following Collection(s)

  • MS [329]

Show simple item record

Search DSpace


Advanced Search

Browse

My Account