Inspiration:
Free stuff!! Not only that but also just being able to learn something new and even delve more into it later on. Also just experiencing something new and getting outside of our comfort zone.
How it works:
It takes unscrambled images as input, where we used numpy to create arrays based on the images, PILLOW to shift the images around, and OpenCV2 to use object detection in the images and find the combination of images with the most confidence.
How we built it:
Started out simple and kept iteratively testing on the training data to improve our program.
Challenges we ran into:
Basically everything and figuring things out, but more specifically when the AI would connect the outide edges with each other instead of the object, especially if it was with white edges.
Accomplishments that we're proud of:
When we matched our first image, as well as earning a score more than 0!
What we learned:
We learned a bit more of how different libraries work and how to use them.
What's next for Unscrambling Image Project:
Continuing to refine it and finding better ways to match images.
Built With
- opencv2
- pillow
- python
Log in or sign up for Devpost to join the conversation.