As I again choose to update my blog instead of the official Indoor Aerial Robotics Competition site, I am pleased to announce that our GP-722 2.4 GHz wireless camera receiver module is properly functioning and outputting video to my computer. This was one of the steps I was most apprehensive about because of several factors:
- The receiver module has almost no documentation associated with it.
- It has flaky drivers that ceased to be updated after Win XP. Questionable ones I haven’t tried exist for Vista.
- It output is designed to be Composite. (The yellow plug on old video game consoles…) From what I understand, it doesn’t encode the video signal in a format as simple and “plug-and-play” as a webcam’s even though it’s connected to my computer via USB.
The plan for my aerial robot is to use either Processing or OpenCV to perform computer vision operations as required by the competition. Expected operations would include blob detection, or line tracking. For these software programs to work, the video from the wireless camera needs to be “digitized.” I found this out after studying Drexel’s material lists from previous competitions. They employed $100 software, as well as TV tuners. Being a stingy miser at heart, and doing some digging in online forums about why wireless cameras wouldn’t show up in Processing, I determined that I ought to install “WinVDig” (version 1.0.1). This does everything the not-free software does… but for free!
I now have a usable video feed being passed through Processing, upon which I can perform all my vision analysis. Remaining challenges include:
- Compensating for static (dropped frames, etc.)
- Ensuring that any interference between the Arduino and the Wireless Camera do not render each system useless.
- Remaining flexible enough in my coding that I can deal with whatever challenge the IARC assigns on short notice…