Community-scale morphometrics provides an remarkable opportunity to quantify the ecological and evolutionary dynamics of taxa in space and time.
AutoMorph is the suite of tools that we have developed to do high-throughput morphometrics on assemblages of objects imaged on a black background. The software (download below) was designed to work with image stacks (each taken at a different focal depth), but also works on individual images as well as photographs.
We developed this approach for extracting 3D shapes from foraminifera imaged with light microscopy, but have also applied it to other fossils (teeth, ostrocods, pteropods, shells) and objects.
Note, our approach current relies on two commercial software packages (Matlab and Zerene Stacker) both of which can be used in ‘trial’ mode if you’d like to try out our image processing routines.
Download software, tutorials, and example images and image products here and on our GitHub site (https://github.com/HullLab):
SoftwareBasic Processing |
|
Software2D Outline & Measures |
|
TutorialWarning: Outdated |
|
TutorialWarning: Outdated |
|
Image Stack |
|
Segment Output |
Example output (four objects) from ‘segment’ of the example image stack. |
Focus Output |
Example output after applying ‘focus’ to the four example objects from segmenting the image stack. |
2D Output |
Example output from ‘run2dmorph’ of the four example images after ‘segment’ and ‘focus’. |
3D Output |
See recent publication of Hsiang et al (2016), for example output and strengths and weaknesses of the method. |
For example, at one North Atlantic site we have imaged the microfossil community in four slides, each accessioned to the Yale Peabody Museum, containing more than 4,000 objects (thumbnails below, click to download full slide):
Each set of slide images was processed through segment, focus, and run2dmorph, for 2D morphometrics and on-going 3D morphometric research. The complete 2D dataset is available here: