Cinematic Bokeh rendering for real scenes
Authors:
Open Access
- 24 November 2015
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 1-10
Abstract
In this work, we present a novel approach for rendering high-quality Bokeh and depth-of-field for real, non-computer graphics scenes by synthesizing a real lens and involving depth. Although this discipline is well-known to methods in computer graphics, their transition to real content is pending due to the lack of real scene geometry. Our approach utilizes a highly integrated RGBD camera as well as a lens Bokeh archive which is derived from high-quality lens point-spread function measurements. Our novel kernel stitching renderer which implicitly employs non-recursive ray tracing, incorporates both contributions. Thereby, we obtain spatially and depth-varying, photo-realistic and high-dynamic Bokeh shapes, as well as inherently realistic depth-of-field renderings per pixel without image-space splatting. This methodology allows for an increased creative choice in the depth-of-field setting without enforcing the final depth-of-field blur being burned-in while capturing. We explicitly show that our rendering methodology compares to real references of a high-quality reference lens.Keywords
This publication has 21 references indexed in Scilit:
- Polynomial Optics: A Construction Kit for Efficient Ray‐Tracing of Lens SystemsComputer Graphics Forum, 2012
- Rendering realistic spectral bokeh due to lens stops and aberrationsThe Visual Computer, 2012
- Real-time lens blur effects and focus controlACM Transactions on Graphics, 2010
- Real-time lens blur effects and focus controlACM Transactions on Graphics, 2010
- Realistic Depth Blur for Images with Range DataLecture Notes in Computer Science, 2009
- Image and depth from a conventional camera with a coded apertureACM Transactions on Graphics, 2007
- Dappled photographyACM Transactions on Graphics, 2007
- Restoring Images Degraded by Spatially Variant BlurSIAM Journal on Scientific Computing, 1998
- An iterative technique for the rectification of observed distributionsThe Astronomical Journal, 1974
- Bayesian-Based Iterative Method of Image Restoration*Journal of the Optical Society of America, 1972