Iris Recognition on Zybo

This Iris recognition tutorial for Zybo will provide instructions on how to modify a consumer camera (webcam) for infrared acquisition and prepare Zybo board for running the Iris recognition app.

Required materials:

Zybo board;

Consumer camera(webcam);

USB hub;

VGA or HDMI monitor;

Keyboard;

Latest Xillinux release from Xillybus;

Remove the webcams front cover to access the lenses.

Unscrew the lens from the sensor and remove the IR filter (red tinted glass) found on the inner part of the lenses.

To enhance image quality two 850 nm Infrared LEDs must be added.

Make 2 holes in the front cover of the webcam and insert the LEDs.

Use an old USB cable for power. Solder the LEDs to the +5v volts and Ground wire of the USB cable. Batteries can be used instead of the USB cable.

Next, we must block all visible light between the eye and the camera. A plastic pipe can't be used for this step. The length of the cover is given by the focal point of the camera after removing the IR filter.

Place the cover on the camera.

Connect the camera to a PC/MAC and run a video app. Test if the image is in focus and well illuminated. Adjust the length of the cover accordingly.

Head to Xillybus and download the boot partition kit and the SD image.

A well documented tutorial is presented here which will guide you through the installation of Linux on the Zybo board.

Once done, insert the SD Card in the board and set JP5 to boot from SD.

Connect the camera and the keyboard/mouse to the UBS hub.

Connect the USB hub to the ZYBO OTG USB port.

Connect the VGA/HDMI monitor.

Connect the Ethernet cable.

Connect the Power Supply.

Next power on the Zybo board and wait for linux to boot up.

Type startx to access the graphical interface.

Run apt-get install libsdl2-dev

If you don't have a package manager installed go to SDL2 website and download the latest source.

Unzip the archive.

tar zxfv SDL2-2.0.3.tar.gz

Run ./configure && make all

Wait for make to finnish and run:

make install

The latest source code can be found on GITHUB.

git clone https://github.com/e-radu/irisRec

make

For checking a person against the database use:

./irisRec_v1 -d /dev/videoX , where X is the id of the connected camera(usually 0)

For adding a person to the database use:

./irisRev_v1 -d /dev/videoX -a "user_name"

Iris Recognition Demo