My name is Leah DeLorenzo, and I am an undergraduate at the University of Massachusetts at Amherst. I am working on an exciting project in the Irschick lab that is part of a broader NSF grant in collaboration with Cameron Siler, Philip Bergmann and Gunter Wagner whose goal is to research the morphology of Brachymeles lizards. Brachymeles is a genus of Skinks within the lizard family Scincidae. Skinks are among the most diverse groups of vertebrates on earth, with over 1,500 known species. They can range in size from about 3 inches to 14 inches long. Brachymeles skinks are distinctive in showing tremendous variation in limb and toe shape and size, with some species being fully limbed, to others being entirely limbless. They also have varying degrees of axial elongation due to the number of vertebrae that they have lost or gained over evolutionary time. Since they are such a disparate group, they are a great organism to use to study evolutionary trends. There are eight species of Brachymeles skinks that I was tasked to 3D model. All originated from the Philippines. We wanted to 3D model these organisms using photogrammetry in order to quantify the evolutionary differences between these species. We chose photogrammetry because it is cost-effective and yields awesome results! All you need is a camera, a stool and good lighting.
The first thing I did was set up a turning stool with LED lights placed around it. I placed each skink in the middle of the stool and snapped a picture. After I took each picture I rotated the stool in order to get shots of all sides of the specimen. I took about 50 pictures of the dorsal side and 50 of the ventral side each skink.
After I took these pictures I loaded them into Photoshop Lightroom. Using lightroom I adjusted the temperature and sharpness settings of each photo to ideal settings for photogrammetry. After post processing, I loaded the pictures for the dorsal side into Reality Capture. The Reality Capture software uses an algorithm to combine 2D images into a 3D model. After the dorsal side of the specimen was complete, I repeated this process for the ventral side.
I combined both the dorsal and ventral meshes of the specimen using Meshmixer. This proved to be a difficult process, but the end product was worth the countless hours of trial and error. I used an additional program called GeoMagic to smooth out some of the rough edges that were byproducts of the combining process.
Here are two examples of the combined models!
Using these 3D models we will perform some quantitative analysis to understand more about the evolution of the body axis in these amazing animals. Through this process, we hope to answer basic questions on how limbless forms have evolved. In January, I will present my research at SICB in San Francisco! In summation, 3D models are awesome and are great tools for researchers!
University of Massachusetts Amherst Class of 2018