BUET_Novochari

https://heroic-daifuku-10a6c4.netlify.app/

https://youtu.be/ljrNaRUzmHI

Astronomic images captured by radio telescopes are often black and white and not quite engaging. Drawing the attention of the distraction-immersed generation of today to sacred branches of knowledge like astronomy gets harder with the obscurity of resources, doesn’t it? Therefore, we felt an urge towards turning these images into enchanting ones which will grab the attention of youths enlightening them in an aura of Astronomy.

Our Solution

What if under the hood of one single web app, the raw Junocam images could be converted into artistic and aesthetic pieces of pictures? And the cherry on top is that, our app ‘Eloquent Juno’ leverages the power of Machine Learning to extract valuable scientific features on the Jovian surface through automated analysis. Additional features include, download, comparison, quizzes, quotes, music videos and the endless scope of sharing the enthusiasm.

How Does It Work?

Once the user enters into the Web-app interface, a beautifully curated, dark themed page is displayed with the options of filtering Junocam images on the basis of title, mission name, date etc. and sorting the images chronologically. Upon clicking the “Apply Filter” button, the app scrapes out relevant images and displays them in the right editor.

The App saliently presents the following features:

1) Basic Editing - Besides changing standard properties of the images namely brightness, contrast, hue, saturation and sharpness, users can add texts upon images, apply color filters and add random shapes. 

2) Scientific Analysis - Leveraging the power of Machine Learning, our app is capable of analyzing Junocam images from a scientific point of view and extracting features on Jupiter surface including The Great Red Spot, Cloud ripples, Swirl Cyclones and Moon Shadow.

3) Quotes - With one tap, you will be able to add interesting randomized or categorized quotes on Junocam images.

4) Background Modification - The app enables you to add custom backgrounds to the images which will be used as alpha masks wrapping the images and turning them more attractive.

5) Comparison of Images - After glamorizing the images with the editor, you’d like to compare it with what it was like before, right? Here comes an elegant slider-based interface to compare the original and edited image.

6) Image Download - You may also save your favorite raw or edited images in your device.

7) Music Lyrics Video - Possibly the most exquisite feature of our app is the automatic generation of Music videos combining scraped music audio and lyrics and edited Junocam images. What could ever be a better way to pique someone’s interest in Jupiter?

8) Scientific Quizzes - Grab the chances of topping the leaderboard by taking part in quizzes based on Junocam images.

9) Amusing Puzzles - Puzzles are fun, right? How'd it feel to solve daunting puzzles generated automatically from Junocam images. Our app scrapes the images, splits them into grid and eventually scatters them to form a puzzle. You can then swap or move the pieces to solve and enjoy.

10) User Manual - We believe, without a well documented user manual, even the most user-friendly application turns into an obscure one at least for the novel visitors. Therefore, we included a detailed user manual on how to use the features conveniently. Click on the right and left arrows buttons to traverse through a detailed guidelines with associated images.

11) Scope of Sharing - Last but not the least, sharing is caring. Spread your utmost enthusiasm about Jupiter among other space-interested fellows through Facebook, Twitter, Telegram and what not.

Tools Used:

• Frontend - ReactJS
• Frontend Deployment - Netlify
• Backend - Python Flask
• Backend Deployment - Heroku
• Image Processing - Python OpenCV, Pillow, Tensorflow

Programming Languages:

• JavaScript
• Python

Link to Codebase:

https://github.com/aaniksahaa/BUET-Novochari_Eloquent-Juno_NASA-Space-Apps-Challenge-2022

Images and Raw Data scraped and sorted from the following sources have been used in the interface.

1. Junocam Images on NASA Website:https://www.missionjuno.swri.edu/junocam/processing?source=junocam
2. Complementary Datasets from Canadian Astronomy Center: https://www.asc-csa.gc.ca/eng/open-data/applications.asp

How We Used the Data:

Our first observation was that the pages where the Junocam images are displayed have a distinguishable processing ID. Then again, the images are being served from a content delivery network which in turn gives the images a distinct vaultID. Upon analyzing these, we used python Regular Expression to scrape the web pages and to extract the image-URLs from the website. Afterwards we used these IDs and associated metadata to filter out the images based on the user’s choice. Furthermore, we used the metadata to train our app how to detect surface features like The Great Redspot, Cloud Ripples, Stroms, Swirls and Cyclones.

Strengthening our interconnections, boosting our confidence and making us aware of our technical lackings, this challenge has served as a blessing for us. Under the kindest supervision of local authorities, we managed to develop and deploy a web app from scratch. Adding one feature after another was a challenging activity indeed. Besides, two of our team members put utmost effort on the presentation and video editing part. The greatest hurdle on the journey was the unbelievable time crunch. Finishing the web app and finalising all features within time seemed quite daunting. And the largest motivation behind the whole hackathon was the urge of representing our country on an international stage. In a nutshell, the journey could not be more memorable. We’ll look forward to such challenges in future.

1. Hansen, C.J., Caplinger, M.A., Ingersoll, A. et al. Junocam: Juno’s Outreach Camera. Space Sci Rev 213, 475–506 (2017). https://doi.org/10.1007/s11214-014-0079-x
2. Orton, G., “JunoCam Imaging of Jupiter Through 28 Perijoves”, vol. 43, 2021.https://ui.adsabs.harvard.edu/abs/2021cosp...43E.465O/abstract
3. The Rich Dynamics of Jupiter's Great Red Spot from JunoCam: Juno Images, A. Sánchez-Lavega et al 2018 AJ 156 162.https://iopscience.iop.org/article/10.3847/1538-3881/aada81/meta

#software, #app, #webapp, #machine-learning, #python, #javascript, #problemsolving, #image-processing