Having recently read a number of posts with regard to the RoboRealm software, I decided to have a look at it. I must say that I’m impressed, and can see how this would make vision processing a lot simpler.
I am, however, finding myself a bit overwhelmed by all of the options, filters, features, and what have you. I’ve learned a lot just by plonking around with the options, building processing chains, reading some online tutorials, and reverse-engineering a bit of code.
With the RoboRealm folks working on adding SSC-32 support for the software, I would very much like to see something for the rank beginners, to introduce the program, and get up to speed on using the software effectively.
What would be extra-cool would be a “Using RoboRealm with the SSC-32” tutorial file that walked the reader through creating a chain that would, for example take the image from a camera mounted on top of a 2-servo pan/tilt platform connected to an SSC-32 via a COM port, analyze it, and produce the proper control commands to allow it to track, say, a blue target object with the pan/tilt rig, to keep the target in the center of the field of view.
I found a feature that does something similar with a robot following a green ball around, but it was more of a “look what you can do” sort of thing, than something that broke down the functions that it was performing to achieve the desired results, and why they were used, as opposed to other similar modules.
For all I know, there may well already be such a tutorial out there, but if so, I’ve missed it.
It cuts everything out but the skin and colors that red. That should make it pretty easy for a robot to see and track a person.
