This scenario assumes a set of images have been captured by one camera at each epoch of an inflating balloon. That is, the balloon is set up with some adhesive targets. Other adhesive targets are placed on a non-moving jig that surrounds the balloon. The balloon is partially inflated so the targets are all visible. A set of photographs is taken as if this were a standard PhotoModeler project.
• The balloon is inflated some more and allowed to stabilize. A new set of photos are taken (using similar camera positions but not exactly the same ones as the first set of photos). We are assuming the balloon does not change shape during the photo shoot.
• The process is repeated at each inflation level.
• The images are placed into sub-folders - one sub-folder per epoch.
• PhotoModeler is started.
• A new project is started using the project wizard, a motion project type selected and then the ‘Type 2 project’ is chosen.
• A camera file (calibration data for the camera used) is chosen.
• The sub-folders with the separated epoch images are identified in the Wizard.
• Once the Wizard is complete, we manually mark, reference and process every epoch. Basically treating each epoch as a separate minimal regular PhotoModeler project. To minimize work we mark and reference only some key points - typically all or some of the static points (those on the jig surrounding the balloon). When done each epoch should have a valid 3d object (some points on the static jig).
• We manually name the static points (4 or more), that are common across the epochs, with common unique names (right click, enter User Name on the 3D Point Properties dialog). We use a naming system for the points on the jig so they are easy to identify.
• Using the ‘Static Points’ dialog we assign at least four static points from the first epoch that appear in or will be tracked to all epochs.
• In the first epoch, we complete the project by marking and referencing all the remaining points (the remaining targets on the balloon). We process the first epoch again to ensure it has a good solution and the balloon a good 3D shape.
• We run tracking with the appropriate options. All 3D points that exist in epoch 1 will be tracked to the specified epochs - including those on the balloon that is changing shape.
• Upon completion of the tracking, manually step through the epochs using the time line control to see if the 3D data makes sense and to ensure tracking was successful.
• We save the project.
• We perform measurements and use the epoch reports.
In a type 1 project one tends to collect all the images from all epochs at once and then set up a big project and track them all. While this can be done with a type 2 project, one can also build the project incrementally. You start with a one-epoch Motion Type 2 project. Then when you capture your next set of photos (there might be hours or days between the epochs in your type 2 project), you append them using the Append Epoch Dialog. You set up the 2nd epoch project and the static points as described and track the remaining points from epoch 1 to epoch 2. This is repeated for each new epoch.