I am a college professor of Physics, have a doctorate in Mechanical Engineering from 1997, and am owner of a scientific instrument company which originally designed and sold astronomy research instrumentation. Besides teaching Physics and Astronomy, using my engineering education and experience, I work as a STEM student research project mentor as part of a five-year $3.1 million federal grant.

One project that I have worked on with students for 3 years is the Scaredrone Project.

Scaredrone (US Trademark: 86077777) is a fully autonomous UAV (Unmanned Aerial Vehicle) drone scarecrow system for deterring birds and other pests from fruit orchards and small acreage farms. There are three parts of the system:

1) The drone quadcopter,

2) The automatic charging landing platform, called "scaredrone PERCH"

3) The stereo vision pest detection system, called "scaredrone EYES"

The drone quadcopter has an on board electronic sonic pest repeller with speaker. It generates bird predator sounds via a pre-loaded software database. The user can retrofit their own quadcopter to work with the Scaredrone system.

The automatic charging landing platform has an embedded computer with an upward looking camera on it that visually locks onto an overhead drone and issues Mavlink commands to the drone to guide the craft in for a precision landing. It makes use of the pattern recognition functions of the OpenCV library of functions developed by Intel. The landing platform automatically recharges the quadcopter. A multi-copter platform called "scaredrone ROOST" is under development for uninterrupted continuous drone farm operations.

The stereo vision pest detection system looks out over a farm and using pattern recognition is able to identify and locate in 3-D space the coordinates of each pest that is present on the field. It then commands the drone to dispatch to that location and begin pest deterring operations. When the stereo vision system determines the threat has been eradicated, the drone is commanded to return for a landing.

Currently trying to migrate software to DroneKit with little success.