Pixelator

An example of our algorithm.

Inspiration

We wanted to make a simple product that sharpens blurry images without a lot of code! This could be used as a preprocessing step for image recognition or a variety of other image processing tasks. It can also be used as a standalone product to enhance old images.

What it does

Our product takes blurry images and makes them more readable. It also improves IBM Watson's visual recognition functionality. See our powerpoint for more information!

How we built it

We used python3 and the IBM Watson library.

Challenges we ran into

Processing images takes a lot of time!

Accomplishments that we're proud of

Our algorithm improves Watson's capabilities by 10% or more!

What we learned

Sometimes, simple is better :)

What's next for Pixelator

We could incorporate our product into an optical character recognition system, or try to incorporate our system as a preprocessing step in a pipeline involving e.g. convolutional neural nets to get even greater accuracy with the cost of higher latency.