Source camera identification for online social network images using texture feature / Nordiana Rahim

Nordiana , Rahim (2018) Source camera identification for online social network images using texture feature / Nordiana Rahim. PhD thesis, University of Malaya.

[img] PDF (The Candidate's Agreement)
Restricted to Repository staff only

Download (168Kb)
    PDF (Thesis PhD)
    Download (2221Kb) | Preview


      The numbers of Online Social Network (OSN) users have grown extensively in the recent decades due to the production of various affordable high technology devices for example smartphones with high-end featured camera, and free online social network apps. However, the rapid growth of this technology has also increased the risk of cybercrime, which exposed users to identity theft, scamming and fraud risk. Therefore, digital image from OSNs may provide authorities with crucial evidence to probe further into the crimes. This highlights the importance of digital image forensic in aiding the authorities to curb the cybercrime issues. Digital Image Forensic is an area of study that mainly focuses on validating the authenticity of digital images by extracting detailed information in those images; including resolution, type of devices, location, times and dates. There are two main methods under digital image forensic, namely source identification which focusing on extracting details of the device used to take digital images, and forgeries detections which focusing on detecting changes made to digital images. This research proposed a technique to identify the source camera of digital image, particularly for OSN images. Images obtained from OSNs web have been processed and modified to meet the OSNs service provider’s requirement prior to publication. The process among others includes reducing its resolutions and size. This process also caused some important information in those images are missing or completely erased, making it difficult or impossible to identify the camera source. In response to this limitation, a new technique was proposed for source camera identification using Gray Level Co-Occurrence Matrix (GLCM) and Gray Level Run Length Matrix (GLRLM) texture feature. Image texture feature is referring to a set of metrics that provides information about the colour arrangement, intensities or selected region of the image which were derived from image processing. Therefore, in this research, image texture features were utilised to propose the new technique which was evaluated based on the percentage of detection accuracy. The results from this research proved that using GLCM and GLRLM features the proposed technique able to identify the source of camera from both original and OSN images with a high detection accuracy of 99.30% and 99.67% respectively.

      Item Type: Thesis (PhD)
      Additional Information: Thesis (PhD) – Faculty of Computer Science & Information Technology, University of Malaya, 2018.
      Uncontrolled Keywords: Digital forensics; Image forensics; Source camera Identification; texture features; Social media
      Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
      Divisions: Faculty of Computer Science & Information Technology
      Depositing User: Mr Mohd Safri Tahir
      Date Deposited: 09 Jun 2021 06:49
      Last Modified: 09 Jun 2021 06:49

      Actions (For repository staff only : Login required)

      View Item