[WIP] Tor Radom LIDAR

Discussion in 'Locations' started by Bienio, Jul 15, 2014.

  1. Bienio

    Bienio Registered

    Joined:
    Nov 18, 2013
    Messages:
    13
    Likes Received:
    0
    Ok then, I will go with 16 quads for the 7,5m width of the track, that makes them about 0,5m x 0,5m.

    EDIT:
    One more screenshot, real vs ALS mesh before and after filtering. As you can see even a really dense point cloud has really a lot of noise, which has to be filtered. Without that you would have a really bumpy surface if you create a mesh too dense. In fact I might make my quads a little bigger, since the filtered mesh isn't perfectly smooth as well.

    View attachment 13919

    EDIT2:

    And finally I managed to conform the track mesh to the point cloud's mesh. This mesh isn't as dense, I'm thinking how dense of a mesh would make sense.
    View attachment 13920
     
    Last edited by a moderator: Aug 14, 2014
  2. Bienio

    Bienio Registered

    Joined:
    Nov 18, 2013
    Messages:
    13
    Likes Received:
    0
  3. Jka

    Jka Member Staff Member

    Joined:
    Jan 31, 2011
    Messages:
    954
    Likes Received:
    213
    Looking good, but try to get rid of those tris and use quads instead.

    Also, you don't need so much polys on those narrow sections.

    Cheers!
     
  4. DurgeDriven

    DurgeDriven Banned

    Joined:
    Mar 20, 2012
    Messages:
    6,320
    Likes Received:
    43
    Looks great :)

    ----


    rF2 needs a funded painters and simmers co-op

    Somewhere structured, that people can help and work with realtime software.

    I think there is potential for this.......end results would be Co-Op funded ISI licences, etc etc

    Just a thought :rolleyes:
     
  5. cyruscloud

    cyruscloud Registered

    Joined:
    Oct 5, 2010
    Messages:
    127
    Likes Received:
    54
    Ok, couple things, first that may be a bit overkill on the width polys pending on width of the track, RF2 is using more but if the track is as narrow as it looks that could be a little overboard. Biggest issue I see is with the pic showing the connection between two sides of the track. The curve on the right has a ton of polys yet still looks like the curve is made in 5 edges or so, in otherwords the polys are wasted on it since it isn't smooth. Also for connecting areas like that between two sections that are an odd shape you really don't need all those polys even if you have Lidar data if that isn't a super important track area and even if it is there is a better way to handle that. Rathen than net it together like that just use a line with snap to vertex on and first draw out your shape and set the points not snapped to the road's basic elevation then add an editable poly modifier to it. After that use the cut tool under editable poly to cut that mesh you have into good quads only using what you need, like in those narrow areas you don't need 5 quads wide being basically funneled down into a point. Its a bit hard to describe but look at some other quality tracks in the viewer and turn on outline mode "o" key and just examine how the mesh is constructed, you may very well be overworking yourself for no real reason.
     
  6. Bienio

    Bienio Registered

    Joined:
    Nov 18, 2013
    Messages:
    13
    Likes Received:
    0
    Cyruscloud, thanks a lot for the tips!

    I have managed to open the track mesh in Dev Mode, that's a big step for me. The mesh still needs some work though.

    Also, is there a place where I could download some 3D objects to put in 3ds max? Like tyre barriers, trees etc. Or does everyone make those by themselves?
     
  7. Led566

    Led566 Registered

    Joined:
    Jan 25, 2012
    Messages:
    358
    Likes Received:
    24
    AFAIK you can reuse the objects contained in the ISI tracks.
    And some Bob Track Builder xpacks are free.
     
  8. Bienio

    Bienio Registered

    Joined:
    Nov 18, 2013
    Messages:
    13
    Likes Received:
    0
    The track is in a somewhat usable state. It's far from finished, but I'm running out of time and this is the state that I will present the track in on my presentation when defending my master's thesis.

    It's not pretty, but the main purpose of this project was to show how aerial LIDAR data can be used to create a track model.

    Thanks to all you guys that helped me!

    Here's a side-by-side comparison with a real lap around the track
    https://www.youtube.com/watch?v=FbkYjMt-b7w&list=UUPVTXbcBXUpG48_mFpfYkIA
     
  9. Rich Goodwin

    Rich Goodwin Registered

    Joined:
    May 3, 2012
    Messages:
    1,219
    Likes Received:
    9
    Awesome dude! Well done!
     
  10. Woodee

    Woodee Registered

    Joined:
    Oct 4, 2010
    Messages:
    4,004
    Likes Received:
    1,058
    Stunning job there!!

    Gosh if only I had the equipment to do this at my local kart track!!
     
  11. GTClub_wajdi

    GTClub_wajdi Registered

    Joined:
    Feb 28, 2012
    Messages:
    3,238
    Likes Received:
    572
  12. kimikaze

    kimikaze Registered

    Joined:
    Feb 17, 2011
    Messages:
    650
    Likes Received:
    23
    Defending your master's thesis is clearly just a mather off formalitiy and time :). Good luck, you have done homework. Great results.
     
  13. jrcn50

    jrcn50 Registered

    Joined:
    Jan 12, 2012
    Messages:
    363
    Likes Received:
    1
    Very impressive!
     
  14. Bienio

    Bienio Registered

    Joined:
    Nov 18, 2013
    Messages:
    13
    Likes Received:
    0
    Hey guys, could anyone proficient in english check if there are any errors in this abstract from my thesis?

    "The main objective of this thesis is to explore the possibilities of using photogrammetric technologies in creation of three-dimensional race track models for use in computer simula-tors. The base for the track creation was data acquired from the national geodetic and carto-graphic resource (orthophoto with a ground sampling distance of 10 cm and a point cloud with a resolution of 12 pts/m2), supplemented with photographies taken on site.
    The data’s accuracy has been assessed by comparing it to GPS RTK measurements. Ortophoto and point cloud’s intensity channel have turned out to be accurate enough for further work, unlike point cloud’s RGB channels, which could not be used because of their poor accuracy.
    Aerial laser scan data is characterized by the occurance of noise, which has been reduced in TerraSolid application. Filtered data has been imported to a specialized computer graphics application 3ds Max. Using orthophoto and point cloud as reference, along with photographs, a 3D model of the race track has been created. During the creation process the main focus was on properly recreating the geometry, especially the track surface. The finished model has been exported and implemented in rFactor 2 racing simulator.
    The data used turned out to be accurate enough to create a geometrically correct 3D model. However, it was necessary to properly filter the ALS data before using it as reference. Furthermore, the relatively low resolution of data is a problem if it is intended to be used in modeling fine details, for example smaller elements of buildings. Modeling the track’s sur-roundings can be based entirely on terrestrial laser scans, but an aerial laser scan is too in-accurate and additional data, e.g. photographs, is required."
     
  15. Guimengo

    Guimengo Guest

    "using photogrammetric technologies in THE creation for use in computer SIMULATORS" -> surroundings and inaccurate, no hyphens needed ;).

    "The accuracy of the data has been" sounds better than The data's accuracy has been... It is always more professional to avoid contractions and these possessive forms.

    "Ortophoto and point cloud’s intensity channel have turned out to be accurate enough for further work, unlike point cloud’s RGB channels, which could not be used because of their poor accuracy." doesn't flow well. Is the intensity channel just for the point cloud? "The intensity channels of both the orthophoto and the point cloud were accurate enough to allow for further work, unlike the point cloud's RGB channels which lacked accuracy."

    Your bring up Aerial Laser Scan Data and a description. Then you go on talking about other methods but the way you started the sentence made it look like the focus would be on aerial data. Besides that, just don't forget to insert all the necessary articles. The and of are important, i.e.: "During the creation process the main focus was on properly recreating the geometry, especially OF the track surface."

    Lastly, you bring up rF2. It's good you say it's a racing simulator, without further info at least the person reading knows the finality of rF2. However you could devote one sentence to your choice of platform (and there mention the ability of utilizing telemetry to confirm the virtual representation matching real life data).
     
  16. Bienio

    Bienio Registered

    Joined:
    Nov 18, 2013
    Messages:
    13
    Likes Received:
    0
    Thanks a lot!

    The intensity channel is just for the point cloud. Orthophoto is a separate product, not related to the point cloud in any way. And it doesn't have an intensity channel (it could be generated, but that's not the point here ;) ).
     

Share This Page