Current MS Students/Devangi Patel
From CSWiki
Galaxy Visualization from Hubble Images
Contents |
[edit] What I am planning to accomplish
[edit] Brief Project Specification
The main goal of "Galaxy Visualization from Hubble Image" is to extract galaxies from a hubble image. After extracting, these galaxies will be visualized using 3D modeling. I will generate a new 3D space model on the basis of the 2D galaxy image extracted.
[edit] High level architecture
The purpose of this system is to detect the galaxies from a hubble Image and these detected galaxies have to be visualized in 3D space.
The approach of my project is :
- Galaxy Cuts
- Use a hubble image as an input which contains number of galaxies
- Change the Image into binary or gray scale using OpenCV
- Find out connected pixels of Image using ConnectedComponent class of OpenCV
- Prepare bounding box around found connected pixels from the image
- Detect Galaxies using Image Segmentation
- Extract Galaxies from a hubble image and make *.tiff Galaxy images
- 3D Modelling and Visualization
- Select a *.tiff galaxy image for 3D visualization
- Visualize Image using OpenGL or Maya software tools
- Render image from 2D planes into 3D space using Billboarding Technique
[edit] Why this is academically interesting
[edit] Intended audience
My intended audience will be astronomers, computer graphics and visualization researchers, students from physics and computer graphics.
[edit] Why this is MS-level work
"Galaxy Visualization from Hubble Images" system will detect galaxies dynamically from a hubble image which is very high level work and it needs lots of research. I will have to do research to detect galaxies; to differenciate them from the transparent region. Dr. Frank Summer an astrophysicist and science visualization specialist in NASA took a hubble Image to detect glaxies by detecting galaxies manually. This system will extend the work by doing it dynamically. Visualization is also a wide area in which I will be researching tools such as Maya and OpenGL for implementation.
[edit] Previous work
[edit] Platforms to be used
Technologies involved in this project :
- OpenCV using C++
- Maya or OpenGL ( using 3D Modeling )
[edit] Anticipated challenge(s)
[edit] Challenges
Technical challenges that I will face are:
- Automatic segmentation / detection of galaxy.
- Transparent mask regions and cut-outs of galaxies.
[edit] Anticipated approach to each challenge
Automatic segmentation or detection is the most challenging task of my project. The anticipated approach for this task is to find out connected pixels from the image and prepare bounding boxes around connected pixels. I will use OpenCV to find out connected pixels and to prepare bounding box. After detection of galaxies, the other challenging task is to find out transparent mask regions and cut-outs of galaxies. I am not very sure how I will extract the galaxy image. I will also have to do more research on rendering 2D images in 3D space.
[edit] What I bring to this work
[edit] My relevant background and experience (CS590)
I have taken computer graphics courses in my undergraduate study. I studied algorithms to draw objects like rectangles, circles, etc. I also learnt to create a 2D animaion using OpenGL. So, it will be easy for me to work with images. Over the last several weeks I have been working on image processing. I also have started to learn OpenCV in the past few weeks, but I will have to do more research on visualization.
[edit] What I find interesting about this work
I am interested in computer graphics & visualization. I have done lots of web programming and this project will further develop my skills in a different area. For me this project is very challeging because I have never done this type of project before. I will have to do lots of reseach to accomplish my goal.
[edit] How this work goes beyond my experience and course work
I have studied conecepts of computer graphics in my undergraduate study, but I have never implement these concepts in real time system. In "Galaxy Visualization from Hubble Image" system, I will implement all these concepts.
