IEEE Consumer Electronics Magazine - September 2018 - 54

Professional Development

Entrepreneurs in
Consumer Electronics
By Tom Wilson

his installment of the "Professional Development" column features an interview with Lukas
Kinigadner, chief executive officer of Anyline.com, founded in 2013 in
the beautiful city of Vienna, Austria.
Tom Wilson: What is Anyline?
Kinigadner: Anyline is the leading
specialist in mobile optical character recognition (OCR). Using a smartphone
camera, Anyline is able to scan and import
all kinds of text, codes, and numbers
with high accuracy and in real time. Our
software development kit (SDK) allows
any developer with an idea to transform
his or her mobile camera into a real-time
OCR tool.
Wilson: How did you get started?
Kinigadner: We had an unconventional start: after university I set up a mobile
app development company with some
friends, and we loved developing but
knew nothing about marketing or sales.
We would develop an app, launch it in the
App Store, and then cross our fingers.
After each failed attempt, we would simply start working on the next app and go
through the process again. We must have
done that literally about 35 times! But it
taught us valuable lessons about how the
App Store works and how important it is
to develop for a concrete need that is relevant to customers. We pivoted to developing apps for other companies that already
knew their customers and how to generate
sales, and we became one of the top three
app agencies in Vienna, Austria.

Digital Object Identifier 10.1109/MCE.2018.2835940
Date of publication: 9 August 2018

54 IEEE Consumer Electronics Magazine

Enabling your grandmother
to scan her electricity meter
with a smartphone requires
handling many more
degrees of freedom.
Wilson: How did this turn into
Anyline?
Kinigadner: A friend of mine was a
cofounder at a start up called mySugr,
which had developed an app to help diabetics monitor and manage their sugar
levels. He explained that users found it
very tedious to manually type in the
blood sugar reading from a standalone
measurement device into their mySugr
app, and we discussed using OCR on
the smartphone to enable a user to scan
in the sugar-level reading directly to the
mySugr app. We were excited about the
potential to improve people's lives, but
it turned out to be a much more difficult
technical problem than it sounded.
We initially estimated a month's work,
and eventually it took nine months!
However, having realized that we had
solved a difficult problem, namely,
OCR on mobile, we realized that there
would be many more companies
whose apps could benefit from OCR
as a part of their product, for whom it
would be more economical to licensein our technology than to develop this
themselves. Thus, we had the foundations for a business and, in 2013, after
gaining high-profile clients, such as Red
Bull, and winning Innovation awards,

september 2018

we had the confidence to focus on
this alone.
Wilson: What was the problem, and
why was this so difficult?
Kinigadner: It turns out that performing OCR on a smartphone is difficult, for several reasons. OCR may have
been solved in the 1980s for scanning
machines, where high-quality cameras
and lenses are a fixed distance from the
text within a controlled lighting environment, but enabling your grandmother to scan her electricity meter with a
smartphone requires handling many
more degrees of freedom. A mobile
environment has so many variables,
starting with the ambient lighting conditions, the distance of the camera from
the target text, its size and font, the
experience level of the user, the low
quality of the camera and lens, and, not
least, the heterogeneity of the cameras.
On the Android platform, there are over
28,000 different camera configurations,
each with their own colors and distortions that affect the image as seen by
the OCR.
Wilson: What is so special about
Anyline's solution?
Kinigadner: Our solution makes it
easy for developers to integrate fast,
accurate, and reliable OCR into their
products, because it is designed to deal
with these camera variables and guide
the user through the process of capturing an image. It can warn the user, e.g.,
that the image is too dark or that he or
she needs to move the phone to avoid a
reflection obscuring the text. These features are only possible because our

http://www.Anyline.com

Table of Contents for the Digital Edition of IEEE Consumer Electronics Magazine - September 2018