Jeremy Wright

Jeremy Wright

Dec 28, 2023

Group 6 Copy 653
0

Tools for Globally Distributed Teams: 3D Scanning

Introduction

When working with teams spread across the globe, there are times when one team member has an off-the-shelf part that another one needs to use as a design reference. This happened to our team recently, when Pen had a camera trap that Jeremy needed to measure for design purposes.

Each different camera trap requires a custom shim that adapts the stereoscope to the camera. Creating the shim is not difficult, but it requires having access to the camera trap to take measurements. However, it is not practical to expect a designer to purchase each new camera trap to take measurements for shims, or for a researcher to ship a camera trap to a designer each time they need a new shim.

Because of this, we decided to try having the researcher (Pen) create a 3D scan of a camera trap and send the resulting file to the designer (Jeremy). We also shipped this first camera trap since we are still figuring out if the scans are going to work well enough.

I have written up the process that we used to process the 3D scan so that it could be used by the designer as part of a larger CAD design. What follows are the technical aspects of that process, followed by some reflection on which aspects can be improved in the future.

3D Scanning Process

Pen 3D scanned the camera trap at his university library, and images of the process can be found here. I can be difficult to access a 3D scanner affordably, and our team is interested in exploring other options including open source photogrammetry using smartphones.

Camera Trap Being Scanned By a 3D Scanner

Processing The Scanned Mesh

MeshLab (free and open source software) was used to decrease the complexity of the STL mesh so that the file size is manageable when used within CAD programs. Without this processing the 3D scan fails to open in CAD software due to running out of memory. The process below can be followed once an STL file is opened in MeshLab.

  1. Click Filters->Remeshing, Simplification and Reconstruction->Simplification: Quadratic Edge Collapse Decimation

Make sure that your filter settings look like the following, but you may need to adjust settings like Target number of faces and Quality threshold for your use case.

MeshLab Simplification Filter Settings

2. Clicking the Apply button will process the mesh, and after a time, the MeshLab viewer should update with the result. Once the progress bar shows that the process is complete, the Close button can be clicked on the Simplification dialog.

3. This new mesh can be exported by clicking File->Export Mesh As, and giving the file a new name. Then the Save button can be clicked to create the new file. I would not recommend saving over the original STL file because some trial and error may be required to get the simplification settings correct for your use case.

MeshLab with Simplified Camera Trap Mesh

Converting the Mesh to a STEP File

This part of the process involves the use of FreeCAD, which is a free and open source CAD application. I used version 0.21.1 for this Lab Note. The goal is to convert the simplified mesh from MeshLab into a STEP file that can be used in a variety of CAD applications. The process to do this is outlined below.

1. Open the simplified STL file that was exported from MeshLab in FreeCAD. To do this, click File->Open and then find and select the STL file. The mesh should now be visible in FreeCAD's 3D viewer.

FreeCAD with STL File Open in 3D Viewer

2. In the Combo View on the left side, there is a document tree that will include the file name of the STL that was opened. Select that object in the tree for the next step.

3. The Part workbench holds the tools required to do the rest of the conversion process, and it has to be made active before proceeding. To do this, click the workbench menu and select Part.

4. Once the Part workbench has been activated, a new menu item will be available in the menu bar. The Part menu item can be seen above the workbench drop down in the screenshot in step #3. Clicking Part-> Create Shape from Mesh will show a dialog box, which can be dismissed by clicking the OK button. The STL mesh will then be processed, and a new shape object will be added to the document tree.

5. The new object that has been created in the document tree needs to be selected for the next conversion step. The name of the object will be a variation of the mesh above it in the tree.

New Shape Object (Based on Mesh Object) in Document Tree

6. With the shape object selected, it can now be converted to a solid that is suitable for exporting to a STEP file manageable size. This is done by clicking Part -> Convert to Solid.

7. Select the new "Solid" item that was created in the document tree.

8. The final export can now be performed by clicking File -> Export to open the export dialog. Set an appropriate name for the file, and select Step with colors as the file type. Click the Save button.

STEP File Type Set and File Named

After clicking the Save button, a second dialog box may appear. If it does, un-check the box labeled Export invisible objects.

Export invisible objects Checkbox Unchecked

After clicking the OK button, the STEP file will be created on disk and should be ready for use in a wide variety of CAD programs.

End Result

The end result of this process was used to design a shim for the Browning HP5 camera trap that Pen scanned. The CAD software used to create the stereoscope is CadQuery. The STEP file that was generated from the 3D scan can be imported by CadQuery within a few seconds, and works well in a CadQuery assembly. Without this simplification process, CadQuery and FreeCAD would both hang when trying to import the resulting file because it was too large (>1GB).

Completed Camera Trap and Stereoscope Assembly

The scan of the camera trap was used to gather measurements and check the fit of the shim in CAD. When 3D printed, the resulting shim fit the camera trap well enough for field use. However, it was difficult to create a shim that fit the trap "perfectly" without doing some test fits and iteration. This could be possibly be due to a lack of experience with this design technique though. As we learn more, we will work to document improvements in this process.

Future Improvements

The main improvement we would like to make to this process is to make it accessible to a broader range of users. The current idea is to use open source photogrammetry with mobile phones to create the 3D scans. An article about this potential solution can be found here. If this process provides models that are dimensionally accurate (at least "good enough"), it could be a great option for makers and researchers around the globe. If we are able to get the photogrammetry method to work well, we will report back.

0 comments

Join the conversation!Sign In

About This Project

Camera traps photos tell you what animals are in an area, but figuring out how many is tricky. E.g. two foxes in a photo doesn't mean there are only two; or a camera might see the same fox twice. But, knowing how many is key to answering questions like "Is the population of a species declining?" We're making a low-cost 3D vision add-on for camera traps. This is because 3D photos include spatial information, and existing mathematical models can use it to make much-needed population estimates.

Blast off!

Browse Other Projects on Experiment

Related Projects

Do Australian bats have what it takes to survive the deadly White Nose Syndrome?

Australian bats are at risk from the deadly fungal disease White Nose Syndrome (WNS), which is expected...

What does whales' poop tell us about the deep ocean ?

Over years, I have built a collection of cetacean fecal samples. While the majority of these samples are...

Probiotics for wild boreal toads facing a deadly fungal disease

A fungal pathogen spread around the globe by human activities now threatens many amphibians, including boreal...

Backer Badge Funded

Add a comment