UrbanWalk

Inspiration

People can access information about landmarks on the internet, but what about the true novelty of these locations? Is it really worth going or do locals think that hour drive might not be worth it? What if there were places in the area that only locals knew about and cherished? UrbanWalk takes the crowdsourced facts and tidbits of information from anyone who has passed by, "captured" the landmark, and added their own thoughts or tips to the profile of the landmark. This can connect visitors to the area and its hidden gems or even let locals passing by learn a bit more about their neighborhood.

What it does

This app recognizes landmarks through the camera and returns the name of the landmark as well as the crowdsourced facts and tips on these local destinations.

How I built it

Google Cloud Platform API, Swift

Challenges I ran into

UI/UX, how do we make this a valuable tool for users? Processing images for landmark recognition was a more complicated process than we expected and even then, the latency was slower than expected.

Accomplishments that I'm proud of

Functional and elegant UI, solid workflow that includes all the minimum functionality we hoped for and can be expanded for many more features.

What I learned

A lot abut Swift syntax, use of Google Cloud API, how important UI/UX is and how thinking as the client can greatly improve the app.

What's next for UrbanWalk

Use of location to improve on landmark recognition. Machine learning for landmarks that are not as well known, but ones that people want to "capture" and contribute to. Reduce processing time for landmark recognition. Switch to OpenCV or other computer vision libraries for cleaner implementation and possible addition of live-feed recognition rather than just "capture." User log-ins as well as ratings and filters for which comments are valuable and thus, presented visible for users. Overlay of comments on top of camera view in relation to landmark position (follows landmark as you move camera).