You are viewing the course site for a past offering of this course. The current offering may be found here.
Lecture 19: Image Sensors (68)
ashvindhawan

I'm a little confused on how this helps us. If we use a birefringent material as described, and a single ray is now incident on 4 separate pixels, won't each of those pixels have a separate color filter, as described by the Bayer pattern on the previous slide? How can we simply "average" the results of the 4 pixels to antialias if they each have a different color filter associated with them?

JefferyYC

@ashvindhawan I think your observation is correct that the 4 neighboring pixels would have different color filters on them. From what I understand, the intuition behind this design is the frequency of change of light goes down from 1x1 box to 2x2 box. So imagine for every original ray per pixel, we shoot it towards 4 pixels instead. We don't need to do any additional averaging I believe: as long as the ray is present in 4 pixels, we achieved filtering out high frequency.

melodysifry

In regards to the question above, I think that the observation you described is actually the point- so that we're able to color filter the same ray with each individual color filter

You must be enrolled in the course to comment