Tuesday, April 29, 2014
My astroimages converted to 3D, a collection of movie clips
I have converted my astronomical images to a various 3D-formats. This time I made a video collection out of my experimental nebula 3D-models.
My astronomical images as an experimental 3D-conversionsThis is a looped 14 min video, click to start and stop. Original movie is in HD720p resolution.
Click the Youtube logo at lower right corner to see this video in Youtube.
Then, please, click the gear symbol, to see the video at 720p HD-format.
3D-stereo images from my astrophotos can be seen HERE
All my astronomical images can be seen HERE
How the 3D-models are done?
Due to enormous distances of cosmic objects, no real parallax can be imaged to get a 3D-information. I have developed a method to turn my images to a 3D-models.
Here is a short and simplified explanation, how 3D images are done:
My 3D-experiments are a mixture of scientific data, deduction and an artistic impression.
I collect distance and other useful scientific information before I do my 3d-conversion.
Usually there are known stars, coursing the ionization, so I can place them at right relative distance. If I know a distance to the nebula, I can finetune distances of the stars so, that right amount of stars are front and behind of the object. I use a "rule of thumb" method for stars, brighter is
closer, but if a real distance is know, I try to use it.
Many shapes can be figured out just by looking carefully the structures in nebula. Like dark nebulae must be at front of the emission ones to show. The general structure of many star forming regions are very same. There is a group of newly born stars, as an open cluster inside of the nebula. The stellar wind from the stars is then blowing the gas away around the cluster and forming a kind of gavitation, a hole, around it. The pillar like formations in the nebula must point to a source of stellar wind, for the same reason.
Since nebulae are practically transparent and the gas itself is emitting light, the thickness of the gas can be estimated by its brightness. Emission of ionized Oxygen, O-III, needs lots of energy. For that reason, Oxygen emission seen in the photo must be at close proximity of the ionizing star(s).
The processing workflow itself is kind of sculpting and the result is always an approximate reality.
I turned the original 2D-image to 3D by using a surface modeling software.
Image is first divided to layers by its content in a image processing software. Each layer is then projected to a 3D-surface. To build the 3D-surfaces and to have a good and realistic forms, I'm using a software, that converts the shades in the image to a 3D-form. I have semi automated the whole process, so it doesn't take too much time to make a 3D-conversion.
The nice thing about the final 3D-model is, that only image elements from the original 2D-image are used!
A typical surface model without textures
Subscribe to:
Post Comments (Atom)
1 comment:
JP I can't believe there aren't more congratulatory comments on your implementation of 3d visualization! Fantastic work and even better results.
Simon
Post a Comment