Mechatronics and Controls lab

Kelly Brandon, Alexa Goldstrom, Zulficar Habib and Jigar Jadav

During week 5, Kelly adapted a K Nearest Neighbors (KNN) algorithm with threshold analysis to improve face recognition and incorporate face learning.

Jigar has gotten the Raspberry Pi (Rpi) setup to run openCV 2.4.2 for Kelly’s computer visions program. He has also configured the Secure Shell (SSH) server and Virtual Network Connection (VNC) server setup on the Rpi to connect it over Wi-Fi from any of the computers on the "MechaNetwork". A webcam has also been setup to capture and display video from the camera through openCV library. Results look promising at 320x240 resolution or below. Anything higher than that produces an upwards of 1 sec. lag in streaming the video.

Alexa worked with both Auto Desk Inventor and Solid Works to model the metal neck parts and assemble them. She also researched other robot necks and the human neck to help determine the final design of CAESAR’s neck.

Zulficar worked with a 3DOF robotic arm to move it vertically and horizontally and get the arm to respond to input commands from a Flexiforce sensor. He also integrated an RSA encryption/decryption software with his under-development Android app. Finally, he successfully configured a Wifly shield to communicate wirelessly with an Arduino board.

Image description

Run-time snapshot of the KNN algorithm implementation