Playing with Epsilon Views

In this assignment, you will use multiple pictures under slightly varying position
to create a large synthetic aperture and multiple-center-of-projection (MCOP) images

You will create
(i) An image with programmable plane of focus
(ii) A see-through effect
(iii) MCOP image

See impressive videos at http://graphics.stanford.edu/papers/shear-warp/

---------------------------------------------------------

(A) Available set
http://www.eecis.udel.edu/~yu/Teaching/toyLF.zip
Use only 16 images along the horizontal translation


(B) Your own data set
Take atleast 16 pictures by translating a camera (push broom)
The forground scene is a flat red colored paper with see through vertical stripes creating a fence
Background scene is a flat book cover or painting with very little red in it
Choose objects with vibrant bright saturated colors

Instead of translating the camera, you may find it easier to translate the scene
Put the digital camera in remote capture time lapse interval mode (5 second interval)

When translating, make sure the spacing between the camera is precisely equal
and the camera moves parallel to the image plane.

Effect 1: Programmable focus
(i) Rebin rays to focus on first plane
(ii) Rebin rays to focus on back plane
(iii) Rebin rays to focus on back plane but rejecting first plane

For simple rebinning at psuedo-depth z, 
	Each column i of targer image is average of column (i + z*k)
Where, i is column number, k is image number (0 to 15) and z is psuedo inverse-depth.
For example, when z is zero, you are focussed at infinity and the targer image is simply
the average of all 16 images.
When z is 1, the first column in target image comes from 
	image 0 column 1, image 1 column 2, image 2 column 3 etc.
But you should use continuous values of z, not just integer z values, otherwise you will notice that column 2 of image 2 will never get used.
Create a sequence of images by varying z from 0 to a reasonable number. Increment of 0.2 should
be sufficient.
When Z is non-integer, do not use nearest-neighbor sample.
You will have to implement a simple filtering scheme i.e. blur the source image before copying
pixels from one of the source image columns.

For rejecting a given plane, instead of taking average of 16 values, take the majority vote.
Median of 16 values will work in some cases, but the most common value will be a more robust choice.
If there is no clear majority, i.e. if the most common count is say below 5, set the pixel to black.


Effect 2: MCOP images
Rebin rays to create a single image with multiple views


Useful links
Synthetic Apeture
http://groups.csail.mit.edu/graphics/pubs/siggraph2000_drlf.pdf
http://graphics.stanford.edu/papers/shear-warp/
http://graphics.stanford.edu/papers/confocal/

MCOP
http://www.cs.unc.edu/~rademach/mcop98.html
http://people.csail.mit.edu/jingyiyu/Multiperspective.html

---------------------------What to submit---------------

	Due Dec 2nd
	All code
	For both datasets:
		Effect 1: 3 focus image results
		Effect 2: Alleast 1 MCOP image
	
	For dataset2: input images

	Send me URL to your assignment. Same page should have link to previous assignments.
	Your subject line should be '[Photo] FirstName LastName Assignment 3, SSN: last four digits'
	

---------------------------End Assignment Description---------------