Advanced Techniques
This article will cover additional techniques that you may use.
Last updated
Was this helpful?
This article will cover additional techniques that you may use.
Last updated
Was this helpful?
NOTE: Agisoft Metashape Professional v.1.6.2 was used in this article.
If alignment is not successful, you may opt to manually placing markers in the photos as an additional set of references for the software to use for alignment.
This is not recommended if the majority of your photos are not aligned as it is a very time-consuming and tedious process, instead this would be viable if there are only a few photos misaligned.
In the Photos panel, select a photo for it to appear in the Model view space, now under a new Photo tab.
With the first photo selected, [Right Click > Add Markers] will add a new marker that can be viewed in the reference panel. Add these markers on to features that are consistent across your photo-set.
Moving on to the rest of your photo-set; [Right Click > Place Markers] to reference in previously created markers. The same feature point should be identified with the same marker so that the software knows how to correlate the photos for alignment. If these feature points change across the photo-set due to obscuring or perspective change, you can always place new points.
You may be using multiple chunks for one of the following reasons:
You have decided to split up your photography into different sections of the subject
Working with the underside of an object/other inaccessible areas during one session of the photography
Photos for a certain part are not aligning
You are redoing a certain part and want to merge it with the existing model
This guide will explain the three types of alignment:
Point-based
Marker-based
Camera-based
While you can use the other types, markers are more effective as you have control on how the software approaches the alignment process. With the Point and Camera-based alignment, the software will have to think for itself which increases processing time and may lead to errors.
Process your set of photos as you normally would for the other chunks:
Ensure that you are working from a new chunk. Right-click in the Workspace to add a new chunk. Double-click on it to assign it as the working chunk. You will only need to produce up to a Dense-Cloud to align chunks successfully.
NOTE: Accidentally adding photos to an already processed chunk (aligned photos and beyond) can irreparably alter the file.
By identifying similar points between the Chunks, the software will be able to align the two clouds together using markers.
Right-click on a part of the model to create a marker:
A minimum of 4 markers is required.
The markers cannot be in a straight line.
Repeat the process on the other Chunk, ensuring the numbered marker is the same across the chunks. If you are using this process to align a top and a bottom of an object, you will need be able to pick out similar features.
[Workflow > Align Chunks] will bring up the menu for Chunk alignment.
Select the Chunks you want to align and change the Method to [Marker-Based].Turn Off Fix Scale, if this is on it will lock the scale of the chunks. Turning it off will ensure it scales itself when aligning.
After alignment, the Chunks are still separate (as if in separate layers), to merge them, use [Workflow > Merge Chunks]
Select the Chunks you want to merge and any properties to carry over.
You will find the merged Chunk as a separate Chunk
Agisoft Metashape provides a set of markers that you can print and use for adding an additional set of feature points for your subject. Refer to the Evaluation Matrix for qualities that may require the use of these additional markers on or around the subject.
[Tools > Markers > Print Markers] can be used to print Agisoft Metashape's native marker system. This will come out as a PDF that you can export and print.
The standard type is 12-bit, if you change these parameters, take a note of this as you will need to input these same parameters during marker detection.
When you are placing markers, please follow these rules:
This needs to be considered during your plan; consider how much of the subject you can see in the frame of your photo, and ensure that there are at least 4 markers visible in every photo.
If you are aligning across the entire subject, then you will need to make sure these are captured in the photos to ensure that the software is using them for alignment.
More will always yield better results.
If these are only being used on the ground, they can be enlarged and weighed down instead.
[Tools > Markers > Detect Markers] is used to automate the marker detection process. Input the same parameters that you used when Printing the printers. If you haven't changed any settings, then the default will be set up already, go ahead and click OK.
You can check out the result of the automatic marker detection by going into the photos:
From here on you can refer back to the basic From Photos to 3D Data Guide to proceed with Alignment:
