Final Year Project, LuminAR with a leap motion

[signalian]

Greetings everyone.
Hope you all are in the best of states. I need a bunch of help from you guys as I am working over a final year engineering project. Its basically innovation to what has already been done at MIT in this video. They basically did it with a camera and I am enthusiastic to do it with Leap motion controller. As I belong to a poor country so any sort of financial support or sponsorship is much highly appreciated.
Anyways I will use following main equipment:
1. Leap motion as camera and depth sensor
2. Servo motor gesture controlled mechanical arm
3. ARM A7 based CubieBoard2(For processing part)
4. AAXA PrJ1 PicoProjector

The time for this project is only one month from today.

[u_cap]

There is no ARM support. There is no ETA for ARM support. You will have to take an existing webcam driver - see here
https://community.leapmotion.com/t/protocole-usb-for-robotics-apps/731

You will have to implement your own feature extraction from the raw images. If you want to try the Leap approach of per-scanline identification of elliptic slices, you can obtain the math for e.g. a least squares approach from the patent filings
https://community.leapmotion.com/t/leap-patent-application-and-other-references/717

Contrast enhancement might be required, see first patent.

To be able to do all this, you will likely have to capture USB traffic from a PC to the Leap, as there is no documentation regarding the Leap-specific host-controller communication (to initialize, select capture mode etc.). Maybe Leap might be willing to assist here?