High-Level Project Summary
Our objective is to create an image editing application that receives images from JunoCam and processes them utilizing different image enhancement techniques. The project consists of four phases: gather JunoCam gallery images, color enhancement, image processing and style transference. Our solution solves the challenge on two fronts: digital image processing, improving clarity and enhancing details; and presenting the information through an artist lens using AI and style transference to provide a new creative vision on the Jovian System. The importance of our project rests in the unfolding of Jovian System's beauty in an attempt to make information more accessible to the general public.
Link to Final Project
Link to Project "Demo"
Detailed Project Description
Currently our application can receive zip files containing images taken by JunoCam and load them into the application data structures combing the RGB color channels and providing a set of functions including: upscaling, sharpening, contrast control and overlay filter along with a function to save the processed image in a PNG file. The application was implemented with Python, using the library OpenCV to manipulate images. Our next goal is to create an interactive interface utilizing Kivy and implement style transferring with Justin Johnson's GitHub repository neural-style in the language Lua.
Space Agency Data
We utilized the JunoCam Image Processing Gallery available in here. The highly detailed images of planet Jupiter and the amount of data induced a lot of appreciation for our Solar System's wonders and the incredible advances made by the scientific community over the last few years. Futhermore, the collaborative and open nature of the Juno Project, with the hundreds of images supplied by amateur astronomers, artists and space enthusiasts alike, provided the motivation to engage in the education and advancement of human knowledge.
Hackathon Journey
Being able to gather with friends to learn about science is always a great time, so we couldn't miss this opportunity. We had a lot of fun exploring Nasa's open datasets for the first time and we fill inclined to dig a little deeper in the future. We learned how to work as a team and have the confidence that with our potentials combined we can reach the stars and dream a little (also, version control is a big bonus). Our interest with this project is motivated by our passion for beautiful space pictures and our desire to learn more about image processing techniques. We solved challenges with empathy and communication, respecting our natural limitations. We would like to thank NASA and the partner space agencies for providing this great opportunity.
References
Tags
#image_processing, #AI, #JunoCam, #OpenCV, #Python