Space Connect| Virtual Planetary Exploration
Virtual Planetary Exploration
Space Connect
Summary
We made a project about Challenge: Virtual Planetary Exploration, in which we aimed to increase accessibility to outer space along with portability. We divided the project into 4 major segments.a) Virtual Solar Systemb) Satellite Image Simulationc) Life Identification using Satellite Imagesd) Quantum Encryption and its Simulations
How We Addressed This Challenge
a) Virtual Solar System
We made a web app that allows one to navigate through space in your browser. It uses three.js to render astronomical bodies.It gives one option to either fly from one planet to another or looks for a planet and observes it or to find its orbit and track its movement. We have rendered and made 3D models of all planets in our solar system.
b) Satellite Image Simulation
We have made a web based GUI to access images and travel through a planet using a simulation created from images that could be used for route identification, research purpose. One can Travel Through out the planet using this on their smart phones and laptops
c) Life Identification using Satellite Images
We are using ML and web scraping to use satellite images of earth and train an ML model to detect plants and water in those photos. This model could be used to detect water and plants on other planets using Images captured by satellites
d) Quantum Encryption and its Simulations
We have successfully completed two quantum encryption protocols simulations and their implementation in python. The reason we insist upon quantum encryption over the modern-day hashing is that it uses spins in addition to existing binary encryption also to hack it one needs to travel back in time.
Usability
It could be used by scientist to study the outer space in real time track activities of satellites, planets, stars on their smart phones. It can be used for tracking satellites position and its route location and using Ai based tech to identify destinations and real time navigation. It can also be used by students interested in space exploration to visualize planets and stars
How We Developed This Project
Virtual Solar System
We made a web app using three.js and angular.js
Satellite Image Simulation
We used python for rendering and segmenting images and used three.js and agular for its implementation
Life Identification using Satellite Images
This process uses Computer vision to find the areas fit for life i.e. areas which have any of water bodies, trees, etc. It mainly consists of 3 parts:
i) Segmentation of the Satellite image:
The satellite image generated by the 1st step undergoes Image segmentation ,which separates all the objects in the image by focussing on edges and boundaries. The image is divided into objects such as the buildings, trees ,water bodies, roads ,barren land etc . Our first algorithm of choice is Mean Shift Algorithm for segmentation.
ii) Finding features for objects using Zernike Moments:
Zernike moments are powerful region-based shape descriptors that are invariant against linear transformations and especially against object boundary deformation. For each of the segments generated by Segmentation algorithm , Zernike moments are calculated and stored.
iii)Classification :
In this process , we need to train a Classifier which can identify the buildings, the trees and most importantly, the water bodies. The Zernike moments used by the above method will be used as features for these segments. The classifier is trained with labels as 'buildings', 'trees', 'water'. After the training , We only need to find the part coming under ‘Life' label.
Quantum Encryption and its Simulations
we made simulation for BB84 protocol and SARG04 protocol using P5.js and implemented its calculation using python.
How We Used Space Agency Data in This Project
We used the open source data available at nasa for redering 3d models of planets and for sample images from satellite for Machine Learning Model
Project Code
Data & Resources
https://climate.nasa.gov/
https://earthobservatory.nasa.gov/
https://solarsystem.nasa.gov/planets/earth/overview/
https://threejs.org/
https://angularjs.org/
https://p5js.org/