Sample Chapter from “Track Match Blend”!

by - 2012/03/21 6 Comments Random

For all of you who are already waiting for  too long for the Track Match Blend DVD here is a little preview chapter about using manual undistortion for a better camera solution and using reference images from Google Maps to check if the solution makes sense.

I am working hard to finish the DVD within the next days! Sorry for the delay!
:)

Here’s the link to footage and blendfile:
download.blender.org/ftp/incoming/libmv/sebastian/using_reference_material.zip

  1. 3pointedit says:
    March 21, 2012 at 12:48 pm

    Good things are worth waiting for. Take the time to do the best you can and we will be very happy, thank you.

  2. Forseti says:
    March 21, 2012 at 2:31 pm

    Very very useful ;-), good job

  3. Christoph Pöhler aka Dracio says:
    March 21, 2012 at 6:13 pm

    Fine and very very usefull.
    enouther good reason to use googel maps :D

    did you DVD contain a Tutorial wich tells you how to make a Tracking probe?
    I made one bye my selft and I am realy not shure if it does his job right….

    To be perfectly honest I think in same Building Centres keepe a Pictue of me hidden under the cash desk wiht the unterline “DON#T ASK, just give hin!!!””
    (If not don’t feel sceard or alarmed I already bouth it ;))

  4. hassan yola says:
    March 21, 2012 at 8:34 pm

    I’ve learn a lot watchink that sample thank sebastian

  5. Tycho says:
    March 22, 2012 at 2:39 pm

    Waaah ! It seems very instructive. I probably buy the dvd as soon as it’d be released.
    But it also give me an idea of improvement that can be added in the camera tracker. Would it be possible to add a “Virtual object / Constraint” in reconstitution mode. I explain :

    * first of all, you create virtual plans or lines and you name it.
    * then you add some constraints between track and virtual object : for instance this trackpoint would belong to the virtual plan “flor”.
    * then you add constraints between the virtual objects like the Gps one (cf http://en.wikipedia.org/wiki/Geometrical_Products_Specification)
    * then the solver use the track data and a min-max solver to try do a perfect track.

    So you can assure that the point in the floor are really coplanar and so better detect if a point is really bad (cause a trackpoint can be good and all the other no so good and the first one gains a big error value thought he was the only accurate).

    Thank for amasssing job you have alredy done

  6. Joeri says:
    March 23, 2012 at 10:28 am

    Looks like that is going to be well worth watching.
    Thanks for your effort.

    The sample chapter looks very good, learned a bunch in only 20 minutes. great.