Image Splicing Localization Using Superpixel and Wavelet Mean Squared Error

Seiga Al Ghifari, Hudan Studiawan

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

Image splicing is a form of image forgery where one image is seamlessly pasted onto another. While image splicing itself is not necessarily illegal and is often used for aesthetic purposes, entertainment, or humor, it can also be misused for criminal activities. To aid digital forensics in detecting the specific location of the spliced image area, an image splicing localization algorithm is required. Previous research has proposed a method for detecting the splicing area by employing superpixel segmentation and noise level estimation. This method involves dividing the photo into smaller parts, estimating the noise level in each part, and clustering these estimated noise levels. By clustering the noise levels, the different clusters represent distinct levels of noise and help identify the splicing area. In an effort to enhance the previous research, this study introduces a new method for detecting the splicing area using superpixel segmentation and wavelet mean squared error. The research demonstrates that the wavelet denoising mean squared error value can be effectively employed to detect the image splicing area, outperforming the noise level estimation method under certain conditions.

Original languageEnglish
Title of host publication2023 International Conference on Information Technology
Subtitle of host publicationCybersecurity Challenges for Sustainable Cities, ICIT 2023 - Proceeding
EditorsKhalid Mohammad Jaber
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages593-598
Number of pages6
ISBN (Electronic)9798350320060
DOIs
Publication statusPublished - 2023
Event11th International Conference on Information Technology, ICIT 2023 - Amman, Jordan
Duration: 9 Aug 202310 Aug 2023

Publication series

Name2023 International Conference on Information Technology: Cybersecurity Challenges for Sustainable Cities, ICIT 2023 - Proceeding

Conference

Conference11th International Conference on Information Technology, ICIT 2023
Country/TerritoryJordan
CityAmman
Period9/08/2310/08/23

Keywords

  • DWT
  • clustering
  • digital forensics
  • image forgery
  • image splicing localization
  • noise level estimation

Fingerprint

Dive into the research topics of 'Image Splicing Localization Using Superpixel and Wavelet Mean Squared Error'. Together they form a unique fingerprint.

Cite this