Inspiration
We had lots of fun watching the squirrels while going to the dining hall, and thought it was really interesting that they, along with lots of the wildlife around the Clemson campus weren't very afraid of humans, so we wanted a way to visualize the data, along with any changes to it.
What it does
Through an iOS application, users can take pictures of animals and upload them, where they are then processed and the data is provided on a map for people to view.
How we built it
We used Unity and Xcode to build the iOS application, and used AWS to store the data, process the images provided using a deep learning model, and provide the data back to the user, where it's then formatted as an easy-to-view map of animal spotting's.
Challenges we ran into
Using Unity and Xcode to develop and build the iOS app posed some challenges as there were a lot of version-specific errors and they ended up eating into our time a lot. Both of us had also never used AWS prior to 12 hours before the hackathon, so learning that skill well enough to implement something was challenging.
Accomplishments that we're proud of
Integrating the AWS .NET mobile sdk well, since we were able to handle requests asynchronously, as well as embed meta-data and do everything rather quickly.
What we learned
We learned a lot about both mobile app development, integration with AWS and using various AWS services, like Amazon Rekognition, using S3 buckets, and implementing AWS Lambda functions.
What's next for Clemson Wildlife Tracker
We would like to improve the data visualization by adding support for viewing heatmaps of the provided data. We also noticed that grayscale images also worked well, so we would like to also add IR cameras running off of Raspberry Pi's for more data in places that students walk in less frequently, as well as when student's aren't traveling there frequently, such as nighttime or during the weekend.
Log in or sign up for Devpost to join the conversation.