ESE 111, Team 26, Professor Stalker

Stalk your professor with ease

Comment

Professor stalker tracks a person across a board.

Professor Stalker started as a raspberry pi project with utilized an arducam pivot setup to track a professor. However, due to OpenCV failing to install onto the pi, it ended being a laptop/arduino project instead.

Our project has been shaped been our issues. A failure with raspberry pi sent us to laptop and arduino. Our lack of a webcam sent us to a phone camera and droid cam. A lack of a phone stand made us recycle old plastic containers and chopsticks into a usable product.

Currently, our project's hardware consists of a laptop, an arduino, a makeshift servo/phone stand, a phone and a servo. All of these parts are integrated together thanks to DroidCam and PySerial, the former converts live phone footage as windows recognizable webcam footage and the latter sends data from python code to running arduino code.

Our software makes use of C++, Python, OpenCV and HAAR Cascade. Our python code utilizes machine learning algorithms and large data bases from openCV libraries and HAAR Cascade for live professor stalking. Our python code recognizes and boxes certain parts of the professor such as professor's face, upper body, or entire body. To optimize this process, these searches are dependent on the failure of the one before. If a face isn't detected, the program searches for an upper body, if that fails, then the entire body. This data is then sent to the arduino, which commands a servo to keep the center of said boxed part within the center of the frame.

Built With

Share this project:

Updates