High-Level Project Summary
We chose to work on a 2 part solution->Developing a system that would rely on community input via an app that would collect photographs and crucial information on the debris and storing it in a database. ->The .exif file of the image will be extracted because it would contain all the important information.->The AI would learn in real-time from the information received in the database and be able to send a signal to nearby collection ships for a more robust and automated cleanup response.->Parallely the AI would be used to classify debris from images or videos feeds obtained from a variety of remote sensing sources (example: satellites and drone feeds) and make the same response.
Link to Project "Demo"
Link to Final Project
Detailed Project Description
We created to grab a software that would collect multiple photos from the web so that we could train the AI. And we used Google‘s teachable machine and open CV to train the AI. Technically speaking the AI can identify in class right the debris to an accuracy of more than 96%. We hope to create a system that would be robust and automated so we wanted to create a special response when the debris is identified which was that a signal would be sent out to the nearest collection vehicles so that the debris is not further away than where it was already and it is clean as soon as it was found.
Space Agency Data
We use the debris water to identify the major places where we would find debris, after which we tried to use planet, since it’s a great image database we could check for debris accumulation at places and we realised that we also needed to map for water current so we tried to use meteomatics. However, it wasn’t any help for us. All of this helped us to create “Global Ocean Watcher”.
Hackathon Journey
The Hackathon was a mesmerising as well as an exhilarating event. Our team didn’t have the greatest of expertise but was willing to learn at every point. When we first read the problem statement we understood the security of it and we first started to research about it. Once we were completed with our research and all the team members were on the same page we started dividing the works so that we could all complete our part in time for the entire project to come to life. We faced A lot of challenges and setbacks, the biggest one was the creation of a machine learning AI. When we were faced by said that we spent time researching about it and finding out a solution. Since the team members didn’t have a lot of expertise we took little help from People to understand how to do what we wanted to do.
References
https://teachablemachine.withgoogle.com/
https://getbootstrap.com/
Tags
#water #oceancleanup
Global Judging
This project has been submitted for consideration during the Judging process.