In this part of the project, I took a 17 x 17 set of images, shifted them to the center image, and took the average. For each image, the shift was determined by the difference between the (x,y) point, and the center point, and then multiplied by an arbitrary alpha parameter to determine where the focusing was.
| Alpha = .1 | Alpha = .05 | Alpha = -.05 | Alpha = -.1 |
|---|---|---|---|
|
|
|
|
Gif of Alpha from -.1 to .1
In this part I looked at how the aperture size affected how a view looked. I did this by keeping a constant alpha, but only keeping the images within a certain radius of the center, meaning that we could focus the image by lowering the radius. This works as a proxy for the aperture size, as averaging more images would have the same effect as increasing the aperture size.
| Alpha = 1.0, Radius = 2 | Alpha = 1.0, Radius = 4 | Alpha = 1.0, Radius = 8 |
|---|---|---|
 |
 |
 |
| Alpha = .1, Radius = 1 | Alpha = .1, Radius = 2 | Alpha = .1, Radius = 4 |
|---|---|---|
 |
 |
 |
This project taught me about how you could mimic the ability for a camera to refocus or adjust aperture size by just averaging images from a lightfield with a certian offset. It was really cool to explore!