*Consequently, their centroid size differs significantly from that of the other specimens. I know this is not an issue because the Procrustes analysis removes size,*
This phrase is being repeated too often in this forum and it is not a correct statement. This is only not an issue if you are not going to deal with size at all, all you do is shape analysis. Most of biological questions involve size at some point. This is actually a huge data management issue, and plagues a lot of the 3D data online. Image someone gives you spreadsheet of thousands of measurements, and some are in inches, some in millimeters, and some in meters, and you don't know which is which. Scale/resolution of 3D data are important and not something that needs to be thought after the fact. If you have access to the physical specimen's of these 3D models, scaling is a simple task: Measure a distance on the model, measure the same distance on the physical specimen, and ratio of those is the linear scaling you need to apply to your coordinate and model. Repeat this for every model. Obviously you will be limited by the error in your measurements, the angle you take them etc... But it is better than being orders of magnitude incorrect. If you do not have access to the physical specimen, it may not be possible to scale them. How are these models generated? If these were from microCT scanning, you might trace the steps back (i.e., data import and why the resolution came off wrong) and see if you can calculate a scaling coefficient. On Monday, December 15, 2025 at 10:01:38 AM UTC-8 Pablo Fisichella wrote: > Dear Morphometers, > > I hope you are doing well. I have a basic question, but I don't know how > to approach it. I digitized 3D landmarks and semilandmarks, but a couple of > my 3D meshes are on a different scale. Consequently, their centroid size > differs significantly from that of the other specimens. I know this is not > an issue because the Procrustes analysis removes size, but I wonder how I > can rescale these 3D surfaces to obtain CS values similar to those of the > other specimens. Could someone recommend a procedure for rescaling the 3D > surfaces or the landmarks themselves? Thank you in advance for any > recommendations. > > All the best, > > > Pablo > -- You received this message because you are subscribed to the Google Groups "Morphmet" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion visit https://groups.google.com/d/msgid/morphmet2/8aa2b808-9162-432e-8c32-16b1160346e9n%40googlegroups.com.
