Xilinx-Innovation-Challenge

Team reverse_biased

AUTOMATED-APPLIANCE SWITCHINGS(HOME AUTOMATION)

This is the code submission and video submission of our proposed project in Xilinx Innovation Challenge during Kshitij,2019 at IIT Kharagpur.

Code Requirements:

1. OpenCV 2 or 3
2. Python 3.6
3. PYNQ-Z2 board - for hardware implementation and the computational unit. Also accelerates many functions of the code hence ensuring better FPS ,thus allowing for real-time application
4. USB WebCam(Logitech used here)

Description:

This project uses OpenCV Python library to take in video-frames and process them , to detect the Hand-Gesture shown by the user counting the number of fingers shown . Based on this , this can be implemented on Hardware to maybe switch ON T.V if the user shows 4 fingers and switch OFF if user shows 4 fingers ; using USB Webcam PYNQ-Z2 board is used for Accelerating the code for faster execution.

Logic used for Image Processing:

Based on a fixed background we perform a background subtraction , blur the image and then find out the Binarised image of hand . Then we perform Contour Detection and apply Convex Hull algorithm to determine the count of fingers.

Procedure:

Execute handgesture.py(on PC) or gesture.py(on Board) and for the first 2 seconds keep the webcam faced towards a fixed background maybe a wall or something. Once the working code has displayed Background Captured! then you can move on to bring your hand in front of the webcam , and the code will process the image displaying the number of fingers shown.

Contributors:

1) Sumegh Roychowdhury

2) Sombit Dey

Video Submission:

Drive Link