Vriksh

Demo Video Link

Never Gonna Give You Up

PPT Link

Never Gonna Let You Down

Table of Contents:

• Introduction
• Implementation
• Tech Stack
• Contributors
• Made at

Introduction

Climate change is a very real problem and almost every country is diligently working on making policies to combat climate change. The simplest yet significant step in fighting climate change is 'planting a tree'. Vriksh is a tool for qualitative analysis of a region to predict tree cover and suggests if plantation is required in the region. It performs image segmentation to predict the tree cover percentage.

Implementation

Vriksh performs satellite image segmentation based on U-Net Deep Learning model. The model is trained to predict the tree cover density from the satellite image of the selected region. The training dataset uses satellite images and its' masked images, dividing the image objects into "trees" and "not-trees". Based on the population density and the predicted tree cover percentage, the selected region can be categorised into three zones: green,orange,red. Green zone signifies that the tree cover density is having a healthy balance with average carbon footprint of human population . Orange signifies that the balance is moderate, which calls for future actions to maintain the balance. Red signifies that there is serious imbalance in the selected region. This calls for immediate action.

Tech Stack

1. Deep Learning (U-Net Model)
2. Google Colab
3. Flutter
4. Flask

Contributors

Team: Kothi Bangle Wale

• Saurabh Yadav
• Aakash Kumar Lakhra
• Devansh Srivastava
• Omkar Prashant Juvatkar

Made at