Loading...
- Type of Document: M.Sc. Thesis
- Language: Farsi
- Document No: 40116 (19)
- University: Sharif University of Technology
- Department: Computer Engineering
- Advisor(s): Kasaei, Shohreh
- Abstract:
- Most broadcast stations insert their logos to claim video content ownership and/or for business purposes.This process reduces picture quality in the logo region and also causes visual discomfort to viewers. Logo detection has a variety of applications; such as automatic TV commercial removal, broadcast surveillance, and video content understanding. Video logos can be considered as a form of visible watermark. Stations may also overlay their logos over the original logo when rebroadcasting a program produced by another station. Also, most comapnies keep a giant archive of their broadcasted programs which are going to be used in their future programs. Here, we have presented an efficient method for precise but non real time logo detection and removal from videos by a video completion/inpainting method. By using an efficient gradient-based method we have first located the logo boundaries and then by averaging frames inside those boundaries, we have extracted a precise logo mask even for logos with complicated shapes. In the logo removal process, we have applied an efficient spatiotemporal video completion method which is based on exemplar-based image completion. The removal method processes individual shots in video sequence separately. Our proposed method can successfully extract a precise mask from opaque and semi-transparent logos. The completion method maintains temporal continuity -which is the main challenge in the video completion process- to a great extent. Extensive experimental results show that our proposed method is capable of removing various types of video logos properly and in most cases indistinguishable to viewers. Quantitative measurements also show that our proposed method outperforms a hybrid method (of image completion and interpolation) in reproducing original no-logo video frames by 1.3 dB in PSNR sense, in average.
- Keywords:
- Logo Detection ; Logo Removal ; Image Completion ; Video Completion
-
محتواي پايان نامه
- view
