Introduction to image formation


So far in this lecture series, we've assumed the presence of an image. We've taken an image from a file. We've taken an image from a camera attached to our computer. We've grabbed an image from the Internet. It's time now to talk about how images are actually formed.

Images are formed within cameras and they are actually formed on the retina at the back of our eye. There's a really fascinating aspect of image formation, we take the three dimensional world that we live in, that's very real to us and we project it into two dimensions. So when we have an image, a photograph or something like that, it's flat, it's only two-dimensional. So we've taken three-dimensional things and transformed it into two-dimensional things.

We only had three dimensions to start with, so we've thrown away one, now we've got two. There's got to be some consequence of that. There are some very interesting consequences of trying to crunch three dimensions down into two.

But there's another really interesting aspect of this. And that is if I look at a photograph of a landscape, it looks three dimensional to me, even though I know it's flat, it's a photograph, I can run my hand over it, it is perfectly flat.  So there's some magic going on in my head, where I'm reconstructing dimensions that don't actually exist and that just comes from my own personal experience of the three dimensional nature of the world; my brain fills in this extra dimension.

So these are the aspects of image formation that I want to talk about in this group of three lectures.  In this lecture we're going to look out how images are formed using pinhole cameras and cameras with lenses and the treatment in this lecture is all going to be geometric; it's about similar triangles.  We're also going to look at a few different types of cameras; we're going to look at fisheye cameras and panoramic cameras and a very new type of camera called a "Light Field Camera".

How is an image formed? The real world has three dimensions but an image has only two: how does this happen and what are the consequences? We can use simple geometry to understand what’s going on.

Professor Peter Corke

Professor of Robotic Vision at QUT and Director of the Australian Centre for Robotic Vision (ACRV). Peter is also a Fellow of the IEEE, a senior Fellow of the Higher Education Academy, and on the editorial board of several robotics research journals.

Rate this lesson


Leave a comment