This project, Stalker 9000, is brought to you by Hamza Ahmed, Saqif Bhuiyan, and Edvins Dreijers. This is their senior design project that involves making a drone that can track human movements. It involves using a camera to find the person in the picture and tracking them using a program written in Python.
“Our mission is to provide a platform that would allow Object Tracking to be implemented onto any drone”
~From the Team
Milestones: Building the Drone Programming on Python Putting it all together
Building the Drone:
Our goal for the hardware aspect was to make it low cost and small enough for the drone to be used indoors. The consensus when making a drone is to make it as light as possible, so that it can fly efficiently without using too much power.
Websites used to purchase parts: Amazon, HobbyKing, Hobbywireless
- Matek Micro PDB w/BEC (5V and 12V)
- Turnigy 1300mAh 4S 45C Lipo Pack
- Turnigy MultiStar 32bit Arm 4 in 1 20A Race Spec ESC 2~4S
- KingKong 2204-2300KV Brushless Motor CCW
- 5×3 Carbon Fiber Propellers 5030
- CC3D Flight Controller
- FPV Camera
- Carbon Fiber Frame
- Flysky FS-i6X 2.4GHz 10CH AFHDS 2A RC Transmitter With X6B i-BUS Receiver
Programming on Python:
Our drone will be using image processing to track the object. The programming language used is Python + OpenCV. We will use the HOG system implemented in OpenCV to track a person.
Putting It All Together:
This is the set-up used for the project. The drone sends the video to the computer where the video feed is processed and the directional input is sent back to the drone from the computer.
So far the team has been able to build the drone, and code the program, but has yet to bring them together. The team will continue to work on it, for more information shoot an email to firstname.lastname@example.org.
Hamza Ahmed is an Electrical Engineering student at NYU Tandon School of Engineering. He has been working on building his own 3D Printer and Pick & Place machine at the MakerSpace. In April 2017, he was the first recipient of the Monthly MakerSpace Mini-Grant and has worked on this project.