2. How It Started…
My mom makes handmade paper
that are totally wood-free.
This is her way of contributing
to the environment.
She makes beautiful, hand-crafted
wedding cards and party invites.
Over the years, she has managed
to carve a thriving business out
of this home-spun hobby.
Whenever she finds good fonts
she begins to frantically search
for a match online. Some of her
clients also bring her fonts that they
like but don’t have a name for.
I have often seen her frustrated
searching for a particular font...
This led to an IDEA.
3. What’s the
Have you ever logged on to Shazam?
It’s an online music search platform, where you insert a piece of music
in the search bar, and from the voice of the artist, tune etc.,
the software recognizes the song.
4. Another example: Google Translator
Here you insert a few alien words and
the search giant correctly identifies the language…
5. That Started a Hunt
For a Font Search Mobile App, where
Designers, Illustrators, Artists, Painters,
Calligraphers, Paper Makers, Software
Developers, Gift-wrappers would find
the font they are desperately
hunting for.
A font recognition app for
Android smart phones and IOS.
6. Existing Apps
There are a few such apps with
issues…APP NAME RATINGS PROBLEM/
REVIEW
STYLISH FONTS
(ANDROID
500 Fonts
FONTS FOR
FLIPFONT 50 #6
FONTS
KEYBOARD (IOS)
NEW COOL
TEXT:
FONTS
No Proper Recognition;
‘Font Not Supported
Develop In Sizes; Misleading
Choices + Editing Features
Not Supporting All The
Mobiles Good Fonts
Cool Fonts; Easy
To Operate
Automatic Transformation
To Cool Font
8. The Technology
Human beings have the innate
ability to identify objects,
words , faces, images etc. The
same way, when an algorithm is
applied to a computer, it’s able to
detects tunes, languages,
typeface etc. Image recognition
Technology is a complex mix of
Artificial Intelligence, Machine
Learning, Data Mining and
Deep Learning.
9. Image Recognition
Image or pattern recognition is a
branch of Machine Learning that
focuses on the recognition of
patterns and regularities in data.
Image recognition systems are in
most instances trained from
labelled "training" data
(Supervised Learning) or there
can be new algorithms written to
discover previously unknown
patterns (Unsupervised Learning).
10. The Problem Statement
The problem in developing the font
search app – FontAlly – is finding
The complete database of all the
fonts that are currently available
in the market in the form of
existing APIs..
Or, if this API is not available,
write a new algorithm.
11. Search for a Reliable Database
With this problem in hand, I approached three vendors for their APIs:
fontsquirrel.com
whatfontis.com
findmyfont.com
12. Back-end Integration
If available, my proposed plan
is to write a back-end code
and integrate this database of
existing library of fonts
with the front-end to
create FontAlly.
13. Proof of Concept Model
1.Capture image 2. Convert image
into data or image
processing library
3. Convert with our
existing set of
font database
4. If match
found show
font name
14. Cost-Benefit Analysis
Quick development
Doubtful accuracy,
depending on
API providers.
Total dependence on API.
The solution will not work
when API is not responding.
Costly development cause of it
will entail recurring payment to
the API licensors.
15. Cost Estimate
Time Estimate
The tentative cost of development + admin panel
= INR 60,000/- + GST (approx.)
20 working days.(After obtaining the API)
16. Developing From Scratch
Writing my own algorithm which can be a long-drawn,
challenging process
We don't have all the fonts libraries in
our databases. So First, we have to
create a database with all the
fonts libraries.
Very Hard to create algorithm
for development.
Accurate result.
One time investment
or payment.