So, you guys all wonder why I am here when I don't post any graphics files.

Well, since I'm very busy I don't get too much time to mess with POV-Ray 
unless it is work related.

Here is an animation that another guy and I recently put together.
It isn't very artistic as we are both engineers and like the technical 
aspects of the project better than the aesthetics :-)

The animation is of a very small area in an underground mine.
Look at features in the data to get a sense of scale of the place.
The ceiling is 50-70 feet high!

The sheer amount of data made this a challenging project because 
iterations were not quick.

If we get more work here it will likely be 100x this amount of data... 
gee I can't wait :-o


Anyway, come on over and have a gander


www.stineconsulting.com/junk/


LAter...   Tom

Post a reply to this message

Hi.

I would like to run the larger of the two animations, but I cannot play 
files with the extension "*.m1v". Could you save it as AVI or MPG and use a 
codec such as the new "MS MPEG4" codec? That would be great.

Greetings,

Sven

Post a reply to this message

"Tom Austin" <taustin> wrote in message news:44746fe4$1@news.povray.org...

> The animation is of a very small area in an underground mine.
> Look at features in the data to get a sense of scale of the place.
> The ceiling is 50-70 feet high!


Wow, this is fascinating.  How was the original data collected?

- How

Post a reply to this message

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Sven Littkowski wrote:
> Hi.
> 
> I would like to run the larger of the two animations, but I cannot play 
> files with the extension "*.m1v". Could you save it as AVI or MPG and use a 
> codec such as the new "MS MPEG4" codec? That would be great.
> 
> Greetings,
> 
	Rename the file to *.mpg and it will run. MPEG1 video will work on
a lot more platforms than MS MPEG4...

		Jerome
- --
+------------------------- Jerome M. BERGER ---------------------+
|    mailto:jeb### [at] freefr      | ICQ:    238062172            |
|    http://jeberger.free.fr/     | Jabber: jeb### [at] jabberfr   |
+---------------------------------+------------------------------+
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.2 (GNU/Linux)

iD8DBQFEdJdwd0kWM4JG3k8RAstvAKCK3XuonP9UBB97dgsKMMjkbStdvwCgpm3G
RhH89mvb2fHYYW1cJW63p+w=
=/lU2
-----END PGP SIGNATURE-----

Post a reply to this message

As per comments, the files have been renamed to .mpg.

Windows Media Player runs them fine on my PC here - I hope the same for you.

LAter... Tom

Post a reply to this message

How Camp wrote:
> "Tom Austin" <taustin> wrote in message news:44746fe4$1@news.povray.org...
> 
>> The animation is of a very small area in an underground mine.
>> Look at features in the data to get a sense of scale of the place.
>> The ceiling is 50-70 feet high!
> 
> 
> Wow, this is fascinating.  How was the original data collected?
> 
> - How
> 

Thanks for the interest....

Using a 3D area laser scanner, separate 'scans' are collected at each 
location.

This data is then 'aligned' together using various methods.

Then the aligned data is exported from the scanning software in vrml 3D 
files (which closely match POV-Ray mesh2 format).

A custom routine makes the vrml into a mesh2.

This is done for *each* scan.

So in this animation you are looking at (9) separate mesh2 objects - one 
for each scan.

If you look closely, small triangles are close to scan locations, and 
large triangles are far away from a scan location.  They overlap nicely 
as the alignment went fairly well.

At this time the data is not 'merged' into one cohesive dataset.
We've tried this in various ways only to fail each time.

If anyone has a suggestion on how to create one mesh, please let me know.


LAter... Tom

Post a reply to this message

"Tom Austin" <taustin> wrote in message news:4474ae2b$1@news.povray.org...

> Using a 3D area laser scanner, separate 'scans' are collected at each 
> location.


Forgive the questions, but:  What sort of laser, how long does a typical 
scan take, and what's the maximum resolution?


> So in this animation you are looking at (9) separate mesh2 objects - one 
> for each scan.

> At this time the data is not 'merged' into one cohesive dataset.
> We've tried this in various ways only to fail each time.

I don't suppose each scan has an absolute position reference?  You must know 
something about their positions relative to each other, obviously, or you 
wouldn't have been able to generate the fly-through like you did.

Warp has dealt with meshes quite a bit, perhaps he's got some ideas for 
joining them together...

- How

Post a reply to this message

Because you requested MS MPEG4, I'm assuming you're using Windows.

Windows Media Player will play M1V files.  M1V is simply MPEG1.  The 
reason its M1V and not MPG is that it is a video only file.

-DJ

Sven Littkowski wrote:
> Hi.
> 
> I would like to run the larger of the two animations, but I cannot play 
> files with the extension "*.m1v". Could you save it as AVI or MPG and use a 
> codec such as the new "MS MPEG4" codec? That would be great.
> 
> Greetings,
> 
> Sven 
> 
>

Post a reply to this message

How Camp wrote:
> "Tom Austin" <taustin> wrote in message news:4474ae2b$1@news.povray.org...
> 
>> Using a 3D area laser scanner, separate 'scans' are collected at each 
>> location.
> 
> 
> Forgive the questions, but:  What sort of laser, how long does a typical 
> scan take, and what's the maximum resolution?
> 

	It's a time of flight laser that can gather about 20,000 data points a 
second.  It scans in a spherical grid (2 angles & a range).  The 
scanning window is about 80 degrees vertical and 340 horizontal.
Each of the scans in this animation is over 2 million data points - each 
raw scan file is over 17Mb.  A scan takes about 6 minutes to complete at 
this resolution.  The data is good to about 1/2 inch.

> 
> I don't suppose each scan has an absolute position reference?  You must know 
> something about their positions relative to each other, obviously, or you 
> wouldn't have been able to generate the fly-through like you did.
> 

We do not know anything about the positions of the scans in relation to 
each other except a hand sketch for rough positioning.  We have some 
software that neatly aligns the data automatically.  All we need to do 
is give it a very rough starting point.

Using survey points, we were able to align the data in a real coordinate 
system.


> Warp has dealt with meshes quite a bit, perhaps he's got some ideas for 
> joining them together...
> 

maybe....


LAter... Tom

Post a reply to this message

"Tom Austin" <taustin> wrote in message news:4475ff6a$1@news.povray.org...

> It's a time of flight laser that can gather about 20,000 data points a 
> second.  It scans in a spherical grid (2 angles & a range).  The scanning 
> window is about 80 degrees vertical and 340 horizontal.
> Each of the scans in this animation is over 2 million data points - each 
> raw scan file is over 17Mb.  A scan takes about 6 minutes to complete at 
> this resolution.  The data is good to about 1/2 inch.


Neat.  What type of laser is it?


> We do not know anything about the positions of the scans in relation to 
> each other except a hand sketch for rough positioning.  We have some 
> software that neatly aligns the data automatically.  All we need to do is 
> give it a very rough starting point.
>
> Using survey points, we were able to align the data in a real coordinate 
> system.


Wow, that's pretty slick.  How long does it take to parse and render a 
single frame?

- How

Post a reply to this message