Photogrammetry tutorial 1: equipment

Taking photos for photogrammetry can be dead simple, or it can be a fiendishly difficult task. It all depends on the specimen you want to get a 3D model of. Here’s a not-so-short guide on how to ensure that your photos will get you a good model. Often, much less effort is required, so please experiment on your own. This part 1 will give you a list of tools you need.

There’s four things you really need:

  • Digital camera with quality lens (and polarizing filter)
  • tripod
  • tiny white stickers and a pen
  • scale bar

Optionals include

  • turntable
  • lights
  • ring flash

Digital camera & lens

You can use your phone or a cheap (or expensive) point&shoot camera. But if you want really good models, there is no way around a really good DSLR camera. I use a Canon EOS 650D with a Canon 18-135 mm lens or a Canon 50 mm prime lens. The lenses are OK, but a better one would sometimes help photogrammetry. Same for the camera: if you can get the funding to get an EOS 1D X or another professional model (Nikons are excellent, too, as are some other brands), go for it! The maximum resolution you can achieve depends, among other thins, directly on the number of megapixels your camera’s sensor has. Also, if the circumstances of your photography are suboptimal (bad lighting comes to mind), a better camera may help.

A key advantage of my 650D over many other cameras is the touch-screen display with Life View – you can see, on the display, exactly what the sensor sees, and you can choose your focus point by touching the screen in the right place. Additionally, the 650D screen can be flipped out and rotated so that it points to wherever I want it to point to. That allows very easy handling of the camera even if it is far above my head, or if it is under a mounted skeleton pointing up. Get a camera with such a display! If it means that the sensor is a bit less good, so be it! The ease with which you can take hundreds of photographs that without a rotatable screen would be a real contortionists act simply outweighs the small difference in megapixels between a top-of-the-line camera and an EOS 650D.

Additionally, a circular polarizing filter can work wonders with specimens behind glass. Examples to come – Edmontonia and Protoceratops at the AMNH worked out much better than they would have without a filter. They cost between 70 and 100 USD, quite cheap considering what they allow you to do.

Tripod

A key factor influencing the quality of your final models is the sharpness of your photos. Blurry photos mean less good alignment and especially a bad surface resolution. A tripod allows you to
a) take blur-free photos under bad lighting conditions, but also
b) take photos with long shutter speeds (also called exposure times) and narrow apertures under any conditions.
The latter point is important, because narrower apertures mean large depth of field. And that means that more of your specimen is in focus.

Thus, for a good model, you want to max your f-stop and thus have to use long exposure times (I often use 30 s). That means high quality lenses and a tripod to keep your camera perfectly still.

You should not try to be very economical when buying a tripod. Nor is it good advise to hunt for the least heavy one. Aluminium tripods, or plastic ones, can be significantly lighter than steel ones, but you will rarely get the size and flexibility as well as durability you really want for serious photogrammetry work in the field or collections all across the world.

The size of the tripod you need depends on the types of objects you want to digitize. Assuming you are like me, and want to be able to produce high quality models or any sort of dinosaur bone as well as of complete sauropod mounts, you need a tripod that can put the camera well above your head. Also, it should be able to put the camera just above the floor. Also, you should be able to flip the camera into any direction you want, which means a highly versatile tripod head.

Overall, a good tripod will cost you roughly $300 or more, but that is money well spent. However, any tripod is better than hand-held shooting!

Stickers

Yes, an odd item to list, but sometimes a great help. Usually, you can’t digitize a specimen completely in one go. Gravity gets in the way and results in the object touching the ground in some places. Thus, you will need to flip it upside down to get photos of the underside. Under ideal conditions (more on that later) you can then simply throw all photos into your software and get a perfect results, but normally this means that you need to run the model as two parts: a first and a second half. These need to be stitched, and putting small markers on the specimen helps you aligning the halves properly.

Scale bar

Well, duh! You do not necessarily need an actual scale bar, but you need to have something in your model that
a) you can easily identify the ends of in your model and
b) you know the exact length of.
No, a bone’s maximum length will not do, because you really can’t ever find the exact two spots you measured the length in between on the digital model. Well, almost never. So get something bit in there (the bigger the less will your percentage error be), like a yardstick or a camera lens cover (they are perfectly round, so the biggest distance you can measure is the correct one).

Turntable

For objects that can be set on a turntable there is no better and more comfortable method than using a turntable. Granted, you may end up having to mask out the background on some images, but in my experience you’ll be twice as fast.

Lights

You may, not really need them, but a 5 to 10 LED light you can use to illuminate specimens from below can be really useful, and is small enough to be taken along.

Ring flash

Another method for getting equal and good lighting onto the object is a rung flash. Not too expensive, and very useful for smaller objects, especially on a turntable.

OK, that about covers it! Anything else you need can be gotten locally at short notice. Next post will address the various set-ups you can use. then comes the camera options.

About Heinrich Mallison

I'm a dinosaur biomech guy
This entry was posted in 3D modeling, Digitizing, How to, photogrammetry. Bookmark the permalink.

13 Responses to Photogrammetry tutorial 1: equipment

  1. Pingback: Photogrammetry tutorial 2: picture taking, general remarks | dinosaurpalaeo

  2. Pingback: Sauroposeidon in perspective–too much perspective | Sauropod Vertebra Picture of the Week

  3. NaveenGoud says:

    thank u

  4. Pingback: Photography and illustration talk, Part 12: Stereo and 3D | Sauropod Vertebra Picture of the Week

  5. Pingback: Photogrammetry tutorial 6: building a model from the photos | dinosaurpalaeo

  6. Pingback: How to Make Your Own 3D « paleoaerie

  7. Pingback: PHOTOGRAMMETRY IN PALEONTOLOGY – A PRACTICAL GUIDE | dinosaurpalaeo

  8. Mike Taylor says:

    I’m reading these in order, so it may be that you return to them later and I’ve not yet seen them: but I could really do with some more thoughts on lights and ring-flash (a concept I’d not previously come across at all).

    • Mike, I am using a ring light whenever possible, as it gives perfect lighting (no shadows, even, known wavelength and thus simple white balance). However, it is not very strong, so it only works if I am close to the object, and the light circle is limited in size, which again means I need to be close.
      A ring flash is usually much stronger, which addresses those two problems, and thus allows a bigger distance. However, as with any flash, exposure control is much more of a bother and makes shooting much slower.

  9. Pingback: Photogrammetry: index to Heinrich Mallison’s tutorials | Sauropod Vertebra Picture of the Week

Leave a comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.