By paying a small handling fee, you can take have access to 1000 images from 50 eyes with 20 images taken from each eye selected from the full Bath Iris Image Database. All images are of size 1280 x 960 and compressed using the JPEG2000 codec at a bit rate of 0.5 bpp. The images may be decompressed using the Kakadu software.
In conjunction with the University of Bath, Smart Sensors Limited has collected a significant database of high quality (1280 x 960 pixel resolution) iris images for use in research and evaluation. Currently the full database comprises 800 people (1600 eyes) with 20 images of each left and right eye.
A deliberate decision has been taken to start off with images of pristine quality, as there is a significant need in the scientific and research community to understand how iris recognition performance is affected by various image degradations including focus blur, motion blur, image compression, occlusions and gaze angle. Many of these degradations can be simulated for the purposes of research under mathematically calculated conditions which are usually unavailable when capturing images from commercial iris cameras.
In addition to the high quality images, large numbers of ‘non-ideal’ images from the same subjects can also be made available on request.
To access samples of images from the Bath Iris Image Database, please click on the links below:
Featured ProjectMIRLIN SDK
fully featured toolkit for building iris recognition engines, enrolment and ...
Bath Iris Image Database
By paying a small handling ...
SMART SENSORS makes iris recognition available on Android
Smart Sensors Limited has ported its MIRLIN Iris Recognition SDK to the Android operating system for Smart Phones so it can offer strong and economic biometric identification for a new ...
Martin George, CEO of Smart Sensors Ltd, interviewed at BCC 2011, Tampa
See the man himself discussing iris recognition applications at BCC 2011, Tampa