My Research

Hello.

I haven't blogged in a while. Ages really. I check it every day for whimsical comments - but most days I am disappointed. My actual reason is for the handy links on our page to all the we comics I like to read.

I occassionally get people asking me what it is that I'm working on for my PhD. Even Claire still seems unsure. I recently wrote a paper for a conference in Taiwan, which was good for me because it forced me to make sure that I understand my stuff so well that I can explain it to other people. I still haven't heard from the organisers of that conference yet - I hope they don't leave to late because it will make it hard to organise my flights.

Speaking of organising flights - I had a trip to Melbourne recently for a Machine Vision training. My flights were really quite expensive for that one since I wasn't confirmed to be attending it until 2 weeks before it started. It didn't really matter to me though, since scholarship was paying for it. Basically the training was all about how to set up a machine vision application: with lighting, and cameras connecting directly to PCs running special software. It was more targeted at production line quality control kind of stuff, so it wasn't all relevant to me. Especially the software, which I will never ever used, yet it occupied half of the training.

After the training I had allowed myself most of a day of free time for sightseeing. I have to say, I didn't really enjoy it. Sure, Melbourne is pretty cool, and the free cable car going around the city centre is useful, but doing it by myself was just kind of lame. I didn't have any plan for what I wanted to see, so kind of just wandered round fairly aimlessly. In fact, highest on my thoughts was "where can I find a nice place for lunch, and then keep the receipt to claim the cost back with my scholarship". I settled on a pretty average pasta place - and forgot to ask for the receipt! The other day there was someone on TV in Melbourne city centre - I think it was Hughesy (Huuuuughesy!) from Rove - and it did feel kind of cool that I recognised where he was.

Lately I've been working with Dale (my superviser) to put together a list of things we need to buy to build our own ranged imaging system down here at Victoria Uni. Did I even mention that that's what my research is in - full-field ranged imaging. The basic components are a camera - usually a very special one that can shutter super fast; a light source - in our case red and infra red laser diodes; and control hardware for creating the timing signals between them, and other general control stuff. My task has been to source the camera, camera lenses, and lens filters (for blocking colours of light that we aren't interested in).

What basically happens is that we light up the scene with our laser diodes, and the light bounces back and into the camera. With a lot of cleverness and details that I am omitting here for simplicity, we can measure the time it takes the light to travel from our light source to the object and back again. Since light travels at a constant speed, finding the distance is just a matter of d = vt. What's more, we do this for every pixel of the camera. So rather than having an image where the data is representing the colour of the pixel, we now have pixels that give us the distance to the object in that pixel. It's pretty fun stuff. Once we have our system set up here at Victoria I invite everyone to have a look - and we can take a 3D photo of you :)

I'll leave you now with a video I made a few months back of me playing around in front of the prototype system we have running at Waikato University. The colours represent depth by the way - I'm not normally quite that psychedelic.