Overview
The purpose of this project was to recreate the findings of Ng et al. By capturing multiple images over a plane orthogal to the optical axis, complex effects can be simulated using operations like shifting and averagingPart 1
For the first part, we employ the effect of depth refocusing by shifting the images and then averaging.
Below, you can see the images produced from using all the grid images and changing the scalar each time. By multipling our shift by a scalar c, we have some control over what part of the image that is being focused on. Notice how as c increases, the focus moves to the closer objects.
c=-1
c=0
c=1
c=2
c=3
radius = 0
radius = 1
radius = 2
radius = 4
radius = 6
radius = 8
radius = 18
Part2: Aperture Adjustment
For this part, we digitally changed the aperture of the image. To do this, we selected which images to average over based on a radius from the center image. We can see that the more images we add, the blurrier the objects in the front become, because they move more in each picture relative to the back objects.
Part3: What I learned
This project helped me better my understanding of light field cameras. It also taught me that I should never underestimate how powerful simple operations can be.