FANR5640/7640: Lab 1 – Ground-based 3D model of Warnell Woodpile (Autodesk ReCap )

Download Lab Data Files Here


3-dimensional surface model from iPhone 6 video (Autodesk ReCap Pro 2018)

In this lab, we will take a look at a ground-based approach of generating a 3 dimensional surface model.  First we will consider the Warnell wood pile example.  However neat you think this output may be, without scale, it is is only a pretty picture – a cool demonstration.  So, your first task is to assign a scale to the woodpile terrain model so we can figure out the volume of our log.  Next, instead of reprocessing my video, you will collect video of an object of your choosing to process.

Part I: Assign a scale to existing woodpile model

  • Open ReCap (Start > All Programs > Autodesk Recap > …) and ‘Load a model’ – load the woodpile2.rcm file.  You should see this:
  • Click on woodpile2.rcm to load the model into the editor
  • Model Settings > Set scene upright  (Model Settings is along the left)
  • Delete everything except the log with the ruler
    • Selection > Lasso  (Selection is along the bottom)
      • Click-and-hold and draw, or
      • Single-click around the areas you do not want and hit enter when you are done
      • Hit the DEL button to remove the information
  • Model Settings > Set scale & units
    • Set units: inches
    • Scale by: Value
    • Click on one end of the ruler, click on the other end, type 36 in the Distance box, hit Set
  • Analyze > Mesh report
    • Volume: approx. 19,400 cubic-inches

Part II: Workflow…

  1. Capture a short video of an object – try to keep it < 1 minute
  2. Frame captures from the iPhone video using VLC
  3. Process in Autodesk ReCap

Frame captures using VLC

You will use the VideoLAN software, VLC, to extract individual frames from out iPhone video.  VLC can be downloaded from their website ( either as a normal installer or as a ZIP package that does not require any installation.  I prefer the ZIP package since all you have to do to is unzip it and run.  I have already downloaded and prepared the VLC package you will use in this course – it is saved in our class folder.  Copy the VLC folder and contents to your working directory and you are ready to do.  I am assuming that your working directory is c:\temp\work1 and that you have copied your iPhone video over to this location.

Command line frame extraction in VLC

Again, I assume 1) your video is saved in c:\temp\work1, 2) you want to save your frames to c:\temp\work1\outframes, and 3) that you have copied the VLC folder into c:\temp\work1\.

  • Open a command line (Start>Run>CMD)
  • Change to your work directory, type the following at the command prompt
  • c:
    cd c:\temp\work1
  • Start the frame extraction.  The name of my video is “IMG_1929.MOV”.  Type the following into a command prompt ..
    Note from classtime.  Ensure you properly source the path to the vlc.exe, your movie, and your output directory (scene-path);  my MOV was recorded at 59 frames/second, so the scene-ratio=118 ensures one frame every two seconds
  • VLC\vlc.exe IMG_1929.MOV --video-filter=scene --vout=dummy --scene-ratio=118 --scene-format=jpg --scene-width=-1 --scene-height=-1 --scene-path=outframes vlc://quit
    •  A VLC instance should start and your outframes folder should begin to be populated with JPEGs
    • –scene-ratio:  my movie file was recorded at 59 frames per second.  The scene ratio of 118 specifies to grab 1 frame every 2 seconds
    • –scene-width and scene-height of -1 indicates to store JPEGs with the native resolution
    • cd c:\temp\work1

Here is another method of extracting frames from a movie using VLC (external site).

Autodesk ReCap

Autodesk ReCap is one of several software packages we will use this semester to convert photographic information into 3D models.  Before diving into ReCap , lets watch this:

  • Load your photos into ReCap
    • You have the option to create Offline or Online in the cloud.  The GIS lab computers meet the offline specifications.  My office and home computer does not meet the minimum requirements, so I have to process them in the cloud.
  • Once the model is completed, Load the model.
  • Set the upright direction, edit all of the unwanted artifacts, set the scale determine the volume just as you did at the beginning of this lab.