论文信息 - Database-assisted Interactive Mobile Image Completion

Database-assisted Interactive Mobile Image Completion

Image completion involves removing unwanted portions of a given photograph. We propose a method of image completion on the Android phone using a small database to search for a matching image. SURF is employed in image matching and image registration. Fast Poisson blending is used to blend the registered database image with the original image.

Henry Wang | Andy L. Lin | A. Lin | Henry Wang

[1] Daniel Cohen-Or,et al. Fragment-based image completion , 2003, ACM Trans. Graph..

[2] Patrick Pérez,et al. Poisson image editing , 2003, ACM Trans. Graph..

[3] Richard Szeliski,et al. Fast Poisson blending using multi-splines , 2011, 2011 IEEE International Conference on Computational Photography (ICCP).

[4] Roberto Cipolla,et al. Hole Filling Through Photomontage , 2005, BMVC.

[5] Christopher Hunt,et al. Notes on the OpenSURF Library , 2009 .

[6] Alexei A. Efros,et al. Scene completion using millions of photographs , 2007, SIGGRAPH 2007.

[7] Edward H. Adelson,et al. The Laplacian Pyramid as a Compact Image Code , 1983, IEEE Trans. Commun..

[8] Jiaya Jia,et al. Image completion with structure propagation , 2005, ACM Trans. Graph..

[9] Internal Report : 2001-V 04 From Lines to Homographies between Uncalibrated Images , 2005 .

[10] Robert C. Bolles,et al. Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography , 1981, CACM.