manual placement of tie point

[tuffi]

hi,

In many traditional photogrammetric software alongside the GCP and CP placement is possible add some tie points manually. Essentially is the same process for adding the GCP but whitout the coordinate inputs. This would be great to resolve sama alignment problems, expecially in some ground photogrammetry situations or where the feature in different positions are very similar.

It's possible to add this feature? Or maybe is already possible?

Thank you

[Alexey Pasumansky]

Hello tuffi,

In the current version of the product you can use marker instances (in Professional edition only) to define the valid tie points between the images. Such markers do not have to be checke on in the Reference pane nor to have any measured coordinates input), only two or more projections defined on the different images. The projections of the markers will be considered as valid tie points projections during camera alignment operation.

[sambagis]

Hey Alexi,
How do we go about doing this process you describe?  I dont see any tutorial information on the topic

[Alexey Pasumansky]

Hello sambagis,

To create the markers that are to be used as valid tie points you need to:
- open the image in Photo view mode by double-clicking on its label in any pane,
- locate the position of "manual tie point", right-click in the location and choose Place Marker -> New Marker option from the context menu,
- open another image where the same point should be marked, right-click in the proper location and choose Place Marker -> <marker label> option from the context menu,
- repeat the previous step for all the images where the manual tie points should be present,
- repeat the same steps for all the manual tie points,
- select all NA (not aligned) cameras in the Workspace pane, right-click on selection and run Align Selected Cameras operation from the context menu (the initial matching via Align Photos command should be already completed and alignment for incorrectly aligned cameras should be reset in advance).

[javajack]

HI I'm rather new to Agisoft Metashape pro. I've read the above and it mostly makes sense, but could someone clarify. Do I start by resetting the non aligned cameras before starting markers? Do I then put the first marker on an non aligned camera or an aligned camera? Once I've placed markers on non aligned cameras do I select them as a group with an aligned camera to re-align, or just the unaligned as a group? If there are two groups of unaligned cameras do I align them as groups. I see that if you make a new marker then all markers after that are just add marker, so do all markers relate to the first marker added? In other programs it is clear there is marker 1, and all photos have tie points to marker 1 until you make a marker 2 and align all th marker 2 together. Thanks