post-processing-27may10-150dpi-med (1)

post-processing-27may10-150dpi-med (1) - Post-processing...

Info iconThis preview shows pages 1–11. Sign up to view the full content.

View Full Document Right Arrow Icon
Post-processing pipeline Marc Levoy Computer Science Department Stanford University CS 178, Spring 2010 Begun 5/27/10. Finished 6/1/10, and recap slides added throughout.
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
! Marc Levoy Camera pixel pipeline ! every camera uses different algorithms ! the processing order may vary ! most of it is proprietary 2 sensor processing: demosaicing, white balancing, sharpening, compression analog to digital conversion (ADC) storage
Background image of page 2
! Marc Levoy Demosaicing (review) ! linear interpolation average of the 4 nearest neighbors of the same color ! cameras typically use more complicated scheme try to avoid interpolating across feature boundaries demosaicing is often combined with denoising, sharpening. .. 3
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
! Marc Levoy Demosaicing errors ! color fringes or moiré ! the cause of color moiré Fne black and white detail in scene is mis-interpreted by interpolation algorithm as color information 4 simplistic 1D color Flter array Fne diagonal B&W stripes
Background image of page 4
! Marc Levoy Common solution: low-pass flter chrominance signal ! color artiFacts are places where chrominance changes abruptly (due to interpolation errors) but only transiently 5
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
! Marc Levoy Common solution: low-pass flter chrominance signal ! color artiFacts are places where chrominance changes abruptly (due to interpolation errors) but only transiently ! use a median flter on chrominance to remove outlier transient chrominance changes [±reeman 1988] replace the chrominance oF each pixel by the median value in a neighborhood this is a non-linear flter 6 5-pixel neighborhood in: out: in: out: spike noise is removed monotonic edges remain unchanged
Background image of page 6
! Marc Levoy Common solution: low-pass flter chrominance signal ! color artiFacts are places where chrominance changes abruptly (due to interpolation errors) but only transiently ! use a median flter on chrominance to remove outlier transient chrominance changes [±reeman 1988] replace the chrominance oF each pixel by the median value in a neighborhood this is a non-linear flter ! summary oF algorithm 1. apply naive interpolation 2. convert to Y’CbCr 3. 4. reconstruct R, G, B From 7 (wikipedia) Y’ Cb Cr
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
! Marc Levoy Comparison ! take-home lesson: 2/3 of your data is made up ! there are better and worse ways to do this 8 linear interpolation median-Fltered interpolation (crop from larger image)
Background image of page 8
! Marc Levoy Camera pixel pipeline 9 sensor processing: demosaicing, white balancing, sharpening, compression analog to digital conversion (ADC) storage
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
! Marc Levoy White balancing (review) ! 1. fnd the color temperature oF the illumination as an (R,G,B) ! 2.
Background image of page 10
Image of page 11
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 99

post-processing-27may10-150dpi-med (1) - Post-processing...

This preview shows document pages 1 - 11. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online