00000ctmaa22000003a 4500 UP-1685675941123958180 Buklod 20180810113201.0 aa rb |||1 u| ta 180810s xx d r |||| u| (iLib)UPTAC-00019010298 UPVTC eng LG 993.5 2018 C66 C67 Cortez, Danne Marie M. author. Twinning copy-move forgery detection using discrete wavelet transform and uniform rotation-invariant local binary pattern Danne Marie M. Cortez; Ryan Rey M. Daga, adviser. 2018. xii, 85 leaves illustrations. Undergraduate thesis (B.S. Computer Science) -- University of the Philippines, Tacloban. Copy-Move is a kind of image manipulation where a part of the image is copied and pasted into another part of the same image. Since images are widely used in legal services and presenting news, copy-move forgery detection in images is beneficial. Many of these detection techniques exist but most fail at detecting clones when tampered images undergo post-processing operations like blurring, noise contamination and rotation. In this study, a detection technique was proposed using Discrete Wavelet Transform (DWT) to reduce the dimension of an image without losing important details, and Uniform Rotation-Invariant Local Binary Pattern (uRLBP) to extract the extract the features. DWT was first applied along the rows and the columns, resulting to a non-overlapping multi-resolution sub-bands. After applying DWT, uRLBBP was applied as a texture operator which describes the spatial information of the image texture according to its circular neighboring points. Feature vectors were compared with each other and feature vectors with high similarities are detected as clones. Detection accuracy of forged pixels in an image, and false positive rate were used to evaluate the performance of the proposed approach. Results showed that proper selection of parameters used can localize the forged region with high accuracy and low false positive rate. Discrete wavelet transform. Image cloning detection. Daga, Ryan Rey M. adviser. FI UP UPTAC UPTAC LG 993.5 2018 C66 C67 Thesis