Use "points" in Meshlab to measure distance between points

[lblovett]

Hello, I am doing comparative statistical tests examining accuracy and precision using traditional tools (callipers), and then with 3D scans (created by DAVID).

I have been trying out many software tools for the "measuring" part, and I have found the free meshlab the easiest to use, and also using the least computer power.

I noticed that meshlab allows users to "select points" (using the button labeled "Pick Points"... and they are "absolutely placed", they have stable positions... what I want to be able to do us use them as "landmarks" which I can place with high precision, and then be able to measure the distances between any two of the points that I have selected.

Alternately, does anyone know of a way to let me use the measuring tool in a more accurate manner? Is it possible to "zoom in and out" while using the measure tool?


Currently it needs me to be zoomed out (meaning my points will not be precise), so I can see *both* points that I want to measure between all on my screen, but if I were able to zoom in to a very precise point on my model, and then place one "end" of the 'measuring tape', and then zoom out, and move over to the "other end point"... then get the distance between the two precise points?
I wish the measuring tools were more flexible (perhaps it is, and I just don't know how to toggle this).

Does anyone else want to be able to do this, or do they know how it might be possible?

Any assistance with this, or suggestions would be greatly appreciated. Thank you in advance.

[hal]

Hello,

I don't know how to use MeshLab for that, sorry, but if you have an other 3D editing software as 3D StudioMax, you can import your mesh, and then draw a line, snapping the vertices on the desired areas.
Then go in Utilities panel and search the Measure tool. There you can find information about your line (the lenght you need to know).

A free alternative is Adobe Acrobat, that support the 3D contents. Pls try to read .

Regards,
Mattia

p.s.: @ Sven: why not to implement in the next David version a ruler that automatically snap to the mesh's vertices the 2 ruler's points and that and that support the zoom an pan view action without quitting the measurement process (eg.: zoom closer, pan the view, place ruler point A, zoom back, pan the view and place point B).

[lblovett]

Thank you very much for the reply, I will continue to check out other tools for this purpose, I worked on this last night and found out how to get a more accurate measurement so I thought I would share the technique I found in meshlab (I have not yet figured out if the "point picker" allows a far more automated "measure between points" tool [it seems too well designed not to have this potential feature]). I wonder if meshlab has potential for use with/as a geometric morphometric package.

Using the button that looks like a "line globe" (see where mouse is pointing in attached image) is the key to placing the two ends of the tape measure accurately.

1)zoom in as close as possible to your mesh, at the location of the first landmark you wish to use.
2) press the "measruement tool" button in the top menu bar.
2.a) click mouse on the point you have zoomed in on. This will place the first end of your tape measure.
3) press the blue coloured "line-globe" button. This interrupts your measurement tool, and allows you to do movements; use standard way of sliding, turning model, zooming out, then double clicking mouse to "center" on the spot on your model where you wish to have the other end of the tape measure land.
3.a) zoom in as close as possible to your mesh (you haven't yet pressed any button other than the blue line globe)
4) now your whole screen should be filled with the "second point for measuring" (the measuring line may appear to be coming from the "wrong direction" this is ok, it isn't).

5) now you press the "tape measure" button in the menu bar again, this "re-activates" the tape measure tool.
6) single click mouse on the point.

Presto- Your much more accurate measurement is provided (if the numbers are obscured by your mesh, you just need to press the "blue line-globe" again, and twist or rotate the model until the numbers are in clear view.

By the way; amazing results so far in both accuracy and precision, DAVID is an impressive tool with much potential for further uses, digital models provide greater freedom and greater precision in placing measures, reducing inter-observer measurement bias. I have a lot of work to go in a short period of time, but will share results as soon as I am done my little project.

[jibberjive]

Thanks for sharing that technique. Any concrete numbers on the accuracy of scanning and digital measurements compared to your empirical measurements?

[dangre]

Iblovett,
Thanks for the info on the MeshLab 'not editing' button. Here is a test I did awhile back on a pair of calipers. Calipers were set at 38.10mm (1.5000in) and MeshLab says it was 38.0662mm (1.4985in). Now, it was hard to pick points in MeshLab and who knows if I picked absolutely perpendicular points but it was close enough for me. And were these surfaces I picked actually 1.5000" in real life? Anyone know how to increase the font size in MeshLab, my old eyes can't read the tiny measurement font very easily. This also makes me wonder how accurate can one make/print their calibration panel? Was the 65mm calibration panel actually printed at 65.0000mm? Maybe I will measure a scan of my calibration panel to see what it looks like.