The Dash Six SDK offers the merging of accuracy and speed, interoperability
and ease of integration for the processing of iris images that comes together to
form a new breed of iris recognition technology.
It does this in three stages: segmentation, encoding and matching. This
process has come about after years of development, engineering and optimisation
in real-word environments.
Segmentation is first step in this process. The image processing defines and
identifies the sclera and the pupil. The iris is then isolated through
segmentation so that a template can be generated. The iris region is represented
in a rectangular or linear image using a process called normalisation. Although
this makes certain features of the eye – such as the eyelashes and eyelid –
appear slightly warped, it allows for a powerful format for the following steps
of iris recognition.
The next step – masking – ensures that environmental factors such as
sunlight, as well as eyelids and eyelashes, are excluded from the normalised
image so that the iris appears isolated. Once the masking is complete, Dash Six
encodes the iris image and generates a template. This template is a binary
representation of the iris image.
Finally, matching occurs, in which two templates are rapidly compared and a
decision is made whether the iris is a match or not.