Inspiration
Deforestation continues to be a significant problem of the today's world. Despite the evidence from the scientific community and news from across the world, it is a difficult issue for the average person to relate to and care about. As such, we built Trees Under AtTech to allow people to see the real effects of deforestation visually at any place they desire. Through this, we hope to spark interest and care towards investing in reforestation efforts.
What it does
Trees Under AtTech is a web page that allows its users to input a location they want to know the deforestation scene of. Using the Google Maps API, the program retrieves an image of the inputted location. This image is inputted into our ML model to get the deforestation value (the health of the forestry in that area). The model was trained on over 40000 google images over time to receive an accuracy 95%.
How we built it
The web page was built using React.js and Flask. The ML model is a Convolutional Neural Network built using TensorFlow. The Google Maps API was implemented into this program using its code written in JavaScript, HTML and CSS.
Challenges we ran into
We wanted to implement the Google Maps APIs to show a pointer on a world map to the location the user inputs, and use Autocomplete with this too. Unfortunately, after 24 hours of perusing the relevant code and documentation, we were unable to link it to our program. One of our group members is a first-time hacker and spent the majority of his time learning to install relevant packages and learning the basics of the languages we were using for this project, which was hard work but, at the end, very rewarding. We also ran into trouble setting up React and Flask for our entire program so that was a learning experience.
Accomplishments that we're proud of
We're proud that, aside these challenges, we were able to overcome them. Our website code compiles and runs successfully. We managed to connect Flask. And after hours to training the model, our model works with 95% accuracy.
What we learned
One of the most important things that we learned was how to connect our front-end running on JavaScript to a back-end programmed with Python. To achieve this, we learnt how to use React in tandem with Flask, an open-source back-end python framework used to develop web systems using python. This proved to be a daunting task. However, through struggling and perseverance, our team learned valuable skills about integrating Python into web applications. Furthermore, our team was formed just an hour before the competition begun and we quickly had to adjust to each others strengths and accommodate for weaknesses so that we could finish the competition on time. Through this, we learned to quickly adapt to a challenging situation and build skills to help us work together as strangers.
What's next for Trees Under AtTech
We plan to update the web page with more features, such as allowing the user to analyze changes in deforestation over a period of time. This would require sourcing past, as well as present, Google Maps data and building a model that could analyze the change in the area over time. Also, we plan to partner with non-profit organizations to bring more awareness to the strong negative impacts that excessive deforestation has on humans and the earth.
Log in or sign up for Devpost to join the conversation.