Projects / PatchMatch: A Randomized Correspondence Algorithm for Structural Image Editing
PatchMatch
Connelly Barnes| Princeton University
Eli Shechtman| Adobe Systems
Adam Finkelstein| Princeton University
Dan B Goldman| Adobe Systems
We have developed interactive image editing tools using a new randomized algorithm for quickly finding approximate nearest neighbor matches between image patches. Previous research in graphics and vision has leveraged such nearest-neighbor searches to provide a variety of high-level digital image editing tools. However, the cost of computing a field of such matches for an entire image has eluded previous efforts to provide interactive performance. Our algorithm offers substantial performance improvements over the previous state of the art (20-100x), enabling its use in interactive editing tools. The key insights driving the algorithm are that some good patch matches can be found via random sampling, and that natural coherence in the imagery allows us to propagate such matches quickly to surrounding areas. We offer theoretical analysis of the convergence properties of the algorithm, as well as empirical and practical evidence for its high quality and performance. This one simple algorithm forms the basis for a variety of tools – image retargeting, completion and reshuffling – that can be used together in the context of a high-level image editing application. Finally, we propose additional intuitive constraints on the synthesis process that offer the user a level of control unavailable in previous methods.
PatchMatch was later generalized and shown to be useful for a variety of computer vision applications like image denoising, object detection, label transfer and more.
PatchMatch is part of the Content Aware Fill feature in Photoshop CS5, and it was also used as a core algorithm in the Video Tapestries, Cosaliency and Regenerative Morphing research projects.
References |
Publications |
Robust patch-based HDR reconstruction of dynamic scenes
Sen, P., Kalantari, N. K., Yaesoubi, M., Darabi, S., Goldman, D. B, and Shechtman, E.
ACM Transactions on Graphics (Proc. SIGGRAPH Asia'12), 31(6), November 2012.
Non-parametric texture transfer using MeshMatch
Chen, X., Funkhouser, T., Goldman, D. B, and Shechtman, E.
In the Adobe Technical Report 2012-2.
Image melding: Combining inconsistent images using patch-based synthesis
Darabi, S., Shechtman, E., Barnes, C., Goldman, D. B, Sen, P.
ACM Transactions on Graphics (Proc. SIGGRAPH'12), 31(4), August 2012.
The PatchMatch randomized matching algorithm for image manipulation
Barnes, C., Goldman, D., Shechtman, E., and Finkelstein, A.
Communications of the ACM, 54 (11), pp 103-110, 2011. Research Highlight.
The generalized PatchMatch correspondence algorithm
Barnes, C., Shechtman, E., Goldman, D., and Finkelstein, A.
In Proc. European Conference on Computer Vision, Crete, 2010.
PatchMatch: A randomized correspondence algorithm for structural image editing
Barnes, C., Shechtman, E., Finkelstein, A., and Goldman, D. B.
ACM Transactions on Graphics (Proc. SIGGRAPH), 28(3), 2009.
