Artifact-free High Dynamic Range Imaging Orazio Gallo UC, Santa Cruz Natasha Gelfand, Wei-Chao Chen, Marius Tico, and Kari Pulli Nokia Research Center, Palo Alto

Avoid any amount of ghosting. Preserve semantic consistency of the final image. Use as much information as possible from the stack.

Ghost Detection with Single Exposure Replacement. [Jacobs et al. '08, Reihnard et al. '05] Ghost Detection with Single Exposure Replacement. [Jacobs et al. '08, Reinhard et al. '05] Collage-like Approaches. [Eden et al. '06, Cho et al. '04] Collage-like Approaches. [Eden et al. '06, Cho et al. '04] Smarter Weighted Average. [Khan et al. '06]

Reference Frame Selection Consistency DetectionHDR GenerationPoisson Blending

Irradiance (E) is the power incident on a surface [W∙m -2 ] Exposure (X) is the radiant energy incident on a surface [J∙m -2 ] X = E∙∆t

5 exposures 4 exposures 3 exposures 2 exposures 1 exposure

a few more results

Future Work Generation of a consistent reference frame. Multiscale approach. Adaptive choice for the shape of the patches.

thank you!

Additional references 1.L. Bogoni. “Extending dynamic range of monochrome and color images through fusion.” International Conference on Pattern Recognition (ICPR), 3, K. Jacobs, C. Loscos, and G.Ward. “Automatic high-dynamic range image generation for dynamic scenes.” IEEE Computer Graphics and Applications, 28(2):84–93, S. B. Kang, M. Uyttendaele, S.Winder, and R. Szeliski. “High dynamic range video.” In SIGGRAPH ’03. 4.E. Khan, A. Akyuz, and E. Reinhard. “Ghost removal in high dynamic range images IEEE International Conference on Image Processing,” pages 2005–2008, Oct E. Reinhard, G. Ward, S. Pattanaik, and P. Debevec. “High Dynamic Range Imaging: Acquisition, Display and Image-Based Lighting.” Morgan Kaufmann Publishers, December P. Pérez, M. Gangnet, and A. Blake. “Poisson image editing”. ACM Trans. Graph. 22, 3, 2003.