Hamp Hack

This weekend was Hampshire Collage's First hackathon, Hamp Hack. It was a 24 hour interdisciplinary hackathon with it's primary focus set on the topics of environmental sustainability, brain and cognition, life hacks and artificial intelligence.



Out team ended up winning an award from GitHub for our project (we each got an awesome Octocat trophy)!

For our project, we built an application that used OpenCV to process whiteboard images from multiple angles to make them more readable. Our hack was inspired by our own experiences in classes with professors that write very fast. An application that could automatically record the text written on the whiteboard and use OCR to make it searchable would be something that would come in quite useful for most people.

Our design centered around a pipeline that processed images from two different camera angles. The first step analysed the images to find the coordinates of the corners. These coordinates were then used to automatically correct the skew of the images caused by the camera angle. The left and right images were then stitched together to create a large panoramic view of the whiteboard. The final step of the process involved using a Gaussian distribution to sharpen the image to make it more readable.

Here is a sample of out before and after images.




This weekend was a lot of fun for everyone, there was a lot of great projects and it is always really cool to see the projects that each team comes up with.