Project Cyclops (OpenCV object tracking robot)

Project Cyclops: Auto targeting robot
UC Berkeley Spring 2010
ME 102B: Mechatronic Design
Arafat Amin: ME
Ben Chu: EECS and ME
Mark Martinez, ME
Mikhail Podust, EECS and ME
Stephen Tu, CS and ME

The 2-axis robot can track arbitrary targets. The images captured from a webcam mounted on the turret is processed with OpenCV. When the object is within a distance threshold, a laser is fired.

The robot can track both colors and faces. The facial recognition software was implemented using a hybrid of the Haar Classifier and the Camshift algorithm. The user is able to remotely select the desired target in real-time through interacting with an intuitive Django web server.

http://www.ocf.berkeley.edu/~benchu/

Category:

Science & Technology

Tags:

• Cal
• UC
• Berkeley
• UCB
• EECS
• EE
• ME
• CS
• 135
• 102
• 102b
• 149
• ME135
• ME102
• ME102B
• EE149
• Electrical
• Mechanical
• Engineering
• Computer
• Science
• Mechatronics
• Robotics
• laser
• webcam
• Open
• CV
• image
• processing
• facial
• recognition
• Haar
• Classifier
• Camshift
• algorithm
• digital
• signal
• signals
• filtering
• PID
• feedback
• control
• robotics
• autonomous
• robot
• robots
• Arduino
• microcontroller
• electronics
• technology
• computer
• science
• electrical
• mechanical
• engineering
• mechatronics
• dynamic
• systems
• gadget
• mechanics
• 2-axis
• turret

[projectcyclopsrpv]

Very nice:)

[n01z3d]

Hey that's great for Killing people automatically! Honestly, cool work, but don't sell your skills to the evil side ...