![]() |
|
![]() |
|
Evans & Sutherland Distinguished Lecture Series Finding the Magic in an Image Stack Michael Cohen Microsoft Research
Host: Claudio Silva Abstract When we look at the pictures we take, we almost always see something we don't like:* Susie looks great, but Grandma has her eyes closed. * The mountain looks good, but why isn't the sky blue? * Eric is in focus, but the background is all fuzzy. * I just couldn't capture the grandeur of the scene (think Canyonlands). * The flash messed up everything. Believe me, it was a very romantic spot!, All of these problems can be overcome by taking multiple pictures of the same scene and adding a dash of clever software (and maybe a bit of your time). In this talk, I will demonstrate some efforts to create that dash of clever software. For those who prefer technospeak - I will talk about: image stacks, panoramic stitching, high dynamic range imaging, graph cuts, Poisson image blending, and mean shift segmentation. Time permitting, I will touch on how we are extending these ideas into video. I will also show you what happens when the artist in me gets a hold of the software. A digital picture uses up about $0.001 of resources. You can store about 1,000 pictures on $1 worth of hard disk or DVD. So why are you still taking only one picture at a time? Take a whole stack!
|
School of Computing 50 S. Central Campus Dr. Rm. 3190 Salt Lake City, UT 84112
801-581-8224 Send comments to webmaster@cs.utah.edu
Disclaimer