M.Sc. Tezi Görüntüleme

Student: Meliha Ece Gürbüz
Supervisor: Assoc. Prof. Dr. Ali GANGAL
Department: Electrical and Electronics Engineering
Institution: Graduate School of Natural and Applied Sciences
University: Karadeniz Technical University, Turkey
Title of the Thesis: Removing Unwanted Objects From Photographes and Video Images by An Exemplar Based Image Inpainting Method
Level: M.Sc.
Acceptance Date: 21/8/2006
Number of Pages: 56
Registration Number: i1755
Summary:

      In this dissertation the subject of removing unwanted objects (example logos, stains, advertisement writings or any kind of images of objects etc.) from photographs and video images without generation any perceivable deformity is handled. For this aim an examplar based image inpainting method that is used for stable images and photographs is reconstructed and developed for video images.

Although the results of removing objects in video images are better than removing objects in stable images same annoying dithers are observed while the flow of video is continuing. To solve this problem a spatial and temporal median filter and a spatial and temporal mean filter are used .

      This dissertation results are compared with other image inpainting methods’ results and especially with the exemplar based image inpainting method which has close algorithm to ours. After the comparison it is observed that our results are more favorable from those obtained by existing other image inpainting techniques.

      Keywords:

Image Inpainting, Exemplar Based Image Inpainting, Erasing Video Logos, Image Enhacement, Spatial and Temporal Filtering, Median Filtering, Mean Filtering