dome dataset processing

How to process dome datasets

this guide was written using screenshots from windows

Before doing this make sure you created a .dome calibration file (see another guide). Now you have captured a set of images from an RTI dome/system:

1. Move the images from the camera card to a folder on your PC (with an ID name)

The card appears as a drive – (safer to make it show filename list not image icons) the quick way for one capture is control A in camera folder, control X. click into destination folder and control V.

2. In relightlab start a new project, it will ask you to select the folder containing your images and show you this Images task view:

3. click on the Lights task on the left:

4. If this is your first use of relightlab click on “Load dome file” and select your dome’s .dome file. It will remember it and in future you only have to use “Select a recent dome”. It will then show the dome calibration information:

5. Now Click on the Crop task and move the crop box using the corner squares (not too tight) – ignore the “Crop images” button unless you want to save a set of cropped images (cropping saves disk space by cropping the originals and produces a cropped output)

6. Now click on the RTI task and you will see your output and algorithm choices:

A good choice for most objects is HSH, RGB, Web, Images as shown above. You can run this several times to produce a standalone .rti file for example (HSH RGB Legacy Images). The Help notes in the program has brief explanations of each choice. Also click on the "Add openlime viewer code" tick-box. 8) Edit the destination using the Directory/File … button to point to a folder where you want the output. Here we have called it hsh.

9. Save your workspace/project now with (top left menu) File->Save Project and put the .relight file into the same folder as your images. This will let you easily load it of you need to make other outputs afterwards.

10. Click on the Export button at the bottom to build your output. You will see the Queue task pane with a progress bar (this takes a couple of minutes, an rbf takes 1min on my desktop). The Queue task will open:

11. Once it has completed you can view your output by clicking on the button (pointed to with red arrow) on the side. It opens in your default web browser:

12. Output other types of data if needed. You can go back to the RTI task and select other outputs. For example HSH or PTM “legacy” format as a .rti file which can be loaded into RTIViewer. RBF can show more details for some objects (e.g. shiny). Different web outputs can be made if your image is very high resolution (Deepzoom and Tarzoom). If you need normals or 3D meshes click on the Normals task which can produce a jpg normals image or a ply polygon mesh (see another how-to document for details).