• DocumentCode
    3313095
  • Title

    Video matting by automatic scribbling using Quadra directional filling of segmented frames

  • Author

    Gupta, Abhinav ; Mangal, Sourabh ; Nagori, Pratik ; Jain, Abhishek ; Khandelwal, Vineet

  • Author_Institution
    Dept. of Electron. & Commun., Jaypee Inst. of Inf. Technol. Univ., Noida, India
  • fYear
    2009
  • fDate
    8-11 Aug. 2009
  • Firstpage
    336
  • Lastpage
    340
  • Abstract
    Video matting is defined as the process of changing the background of the video by extracting the foreground image from each frame of a video. In this paper, we propose a video matting technique which uses Ncut segmentation technique to segment the frames. Further, Sobel edge detection technique is used on the segmented frames for detecting edges. Afterwards, Edge thickening process fills up all the gaps in between the edges and Quadra directional filling method generates the probability maps (PMs). The frames are scribbled with reference to the constructed PMs. Scribbled frame and the original frame along with a new background image are then fed to closed form matting technique which gives the frame with the changed background.
  • Keywords
    edge detection; feature extraction; filling; image segmentation; probability; video signal processing; Ncut segmentation technique; Quadra directional filling method; Sobel edge detection technique; automatic scribbling; edge detection; foreground image; frame segmentation; probability maps; video extraction; video matting; Bayesian methods; Cameras; Closed-form solution; Data mining; Filling; Image edge detection; Image segmentation; Information technology; Motion pictures; Statistics; Automated Scribbling; Matte; Quadra Directional filling; Video Matting;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computer Science and Information Technology, 2009. ICCSIT 2009. 2nd IEEE International Conference on
  • Conference_Location
    Beijing
  • Print_ISBN
    978-1-4244-4519-6
  • Electronic_ISBN
    978-1-4244-4520-2
  • Type

    conf

  • DOI
    10.1109/ICCSIT.2009.5234624
  • Filename
    5234624