This tutorial is for JAG3D version 3.3.20110211 (published Feb. 11, 2011)
***The interface in newer versions was completely redesigned, though these instructions can still be useful.
The first hurdle to get over is that the coordinate transformation module in this version is all in German, so be patient-I'll try to translate where I can.
The first thing to do is set up JAD3D to run a little more efficiently. In the extracted ZIP file there is a JAD3D_en.bat file that we'll need to edit.
- Edit the file by right-clicking and choosing ‘Edit’ from the context menu.
- We need to increase the amount of memory JAG3D can access through the Java Virtual Machine.
- In the batch file the variable -Xmx controls this.
- Save and Close the BAT file.
- Change the number after this tag to the maximum memory you wish to
use, in Megabytes (1000Mb = 1Gb). I usually put it up to just under the
maximum RAM for the computer.
To run JAG3D with these new settings, simply double-click the JAG3D_en.bat file to launch the program. This also leaves open the Command Line window to view the progress of the transformation.
- If you get a error “Error occurred during initialization of VM…,” your memory allocation is not right (usually too much). Go back and edit the BAT file with a smaller memory parameter.
To prep your data for JAG3D please see my other post Data Prep for JAG 3D.
The first step is to import your two datasets.
- Under the ‘Datei’ menu, select ‘Import Startsystem’, navigate to the
text file with your raw data. Your data will populate the Statsystem
table when it is complete.
Depending on the number of points and the memory you allocated, this could take upwards of 10 mins
- Go back to the ‘Datei’ menu and select ‘Import Zeilsystem’, navigate to the text file with your GCP data. Switch to the 'Zeilsystem' tab on the bottom of the window to ensure your points loaded correctly.
- Choose the '3D' option, in the drop down choose the
7-Parametertransformation-3D (M, Rx,Ry,Rz,Tx,Ty,Tz) option. This will scale the points uniformly and calculate independent coefficients for rotation and translation in all three axes.
Transformationsarten = Transformation types
- The three options for parameter estimation are Unweighted or using
the Gauss-Markov or Gauss-Helmert models of least squares estimation.
I've compared the three estimation options on two different datasets and found zero difference in the outputs. Unweighted in fine for most applications.
Parameterschätzung = Parameter Estimation
If your command prompt window is still open you can watch the progress and it calculates the parameters and applies the transformation. A word of warning, most of the point clouds I've dealt with are between 50,000 and 100,000 points. I've allocated 3.5 Gb to JAG and it routinely takes 20 minutes to process all of the points.
A pop-up will inform you when everything is done. Switch to the Transformation tab on the bottom of the window to see the results.
The next thing to do is export the transformed data.
- In the ‘Datei’ menu, select ‘Export Transformation.’ By default the file doesn’t have an extension. When you type your file name add .txt to the end.
- Save the file to the directory of your choosing.
- This again takes a while depending on the size of your cloud.
- Select all the text in the command prompt window and copy it to a blank text file.
- In the ‘Datei’ menu, select ‘Report erstellen’. This will generate a HTML report of the outputs.
Report erstellen = Create report