I just took delivery of a DJI Phantom 2 Vision quadracopter and I am very excited by my initial impressions. In a past post I showed a modern computer flight simulation program and the hilarious difficulties that I had flying it from a Radio Controlled (RC) perspective. I am happy to say that with a lot of practice using an Heli-X demo with an actual DJI phantom and using their training balloon mode I have become much more efficient at not crashing. Here is a pic of the real setup:
The cool thing is that there is a first person perspective available via my iPhone through the onboard video camera. The autopilot capabilities should allow a single pilot to compose images and simultaneously fly the aircraft. This is potentially a huge advantage over much more expensive systems that require two operators, one flying and the other working the imaging downlinks.
The downside here is the limited payload and the, unfortunately, poor performance of the stock video camera. I am actually having some regrets that I did not get the GoPro setup with the gimbal mount. Even though I’m not a big fan of the GoPro I think that I may have been able to modify the setup to mount my Nikon N1 J1 camera with it.
Here are some comparisons of the stock Vision camera versus a cheap Nikon Coolpix S220, a Nikon N1 J1, and a Nikon D800. First the stock Vision camera:
Yikes, at 900 ISO this is not very good, auto exposure or color rendition. To be fair I did not expect professional results from this relatively inexpensive setup and first person perspective will not be available with the next two setups. The next image is from the cheap point and shoot Nikon Coolpix S220:
Better color and less noise as well as obviously better resolution. Next is the Nikon N1 J1:
Even better resolution and color, this camera would add 3/4 pounds to the load and I’m not sure if the Phantom will handle it well, at least in the stock configuration. Here is an image from a D800 for comparison, the D800 would never be able to be flown by the Phantom, but it is interesting to see what a much more expensive platform would be able to offer:
Note that the double banding down the center of the guitar that is not resolvable by the previous three cameras. Advantage is definitely with platforms that can handle much higher loads, something to shoot for in a future model. For now I’m just getting my feet wet with a relatively cheap platform, and believe me I will be taking this thing apart and analyzing it with the expectations of designing a much more powerful design that can handle a professional camera and fly as easily as the Phantom 2 Vision…
Stay tuned this is just the beginning of something that should turn out to be both challenging and fun, time to blow the dust off of my signals and systems and my control systems textbooks as well as folding in my BLDC motor control algorithms that I’ve been developing…