INTLSnip

Inspiration

I've always wanted to build a snipping tool that automatically adjusts its content to scale. I was inspired while I was preprocessing images for DL data. I realized that cropping an image with the built-in snipping tool can cut out some of the important characteristics in an image. However, sometimes we may not want an large image. In order to find a balance between both world, I did some research and learned about the seam-carving algorithm.

What it does

The application is a snipping tool that crops a selected area of the screen. After selecting the area, it process it with the seam-carving algorithm with the desired configuration.

How I built it

I built the GUI with PyQt5 I used numpy, scipy for the implementation of seam-carving algorithm.

Challenges I ran into

I am working on my own, it is quite hard to code everything on your own within 30 hours

Accomplishments that I'm proud of

Getting the GUI to work, I've never used Qt before As well as watching the algorithm does its thing.

What I learned

How to implement a GUI in PyQt5, make a snipping tool. seam carving algorithm

What's next for INTLSnip

It is possible that instead of cropping an image, it captures a short clip of gif. The seam carving algorithm can be apply to videos as well, perhaps I can scale the gif

I can potentially run the backend on cloud and port it to mobile devices.

Potential uses for data augmentation for deep learning