Detection of Duplicated Regions in Images Using 1D-Fourier Transform


21st International Conference on Systems, Signals and Image Processing (IWSSIP), Dubrovnik, Croatia, 12 - 14 May 2014, pp.171-174 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • City: Dubrovnik
  • Country: Croatia
  • Page Numbers: pp.171-174
  • Keywords: Copy move forgery, Fourier transform, Gaussian Blurring, COPY-MOVE FORGERY
  • Karadeniz Technical University Affiliated: Yes


Large number of digital images and videos are acquired, stored, processed and shared nowadays. High quality imaging hardware and low cost, user friendly image editing software make digital mediums vulnerable to modifications. One of the most popular image modification techniques is copy move forgery. This tampering technique copies part of an image and pastes it into another part on the same image to conceal or to replicate some part of the image. Researchers proposed many techniques to detect copy move forged regions of images recently. These methods divide image into overlapping blocks and extract features to determine similarity among group of blocks. Selection of the feature extraction algorithm plays an important role on the accuracy of detection methods. Column averages of ID-FT of rows is used to extract features from overlapping blocks on the image. Blocks are transformed into frequency domain using IDFT of the rows and average values of the transformed columns form feature vectors. Similarity of feature vectors indicates possible forged regions. Results show that the proposed method can detect copy pasted regions with higher accuracy compared to similar works reported in the literature. The method is also more resistant against the Gaussian blurring or JPEG compression attacks as shown in the results.