Free software classifies leaf veins

GEORGIA TECH (US) — A new software tool takes an image of a leaf, analyzes it, and returns detailed information on the structure of that leaf’s vein networks.

The program, called Leaf Extraction and Analysis Framework Graphical User Interface (LEAF GUI), enables scientists and breeders to measure the properties of thousands of veins much more quickly than manual image analysis tools. Plant biologists study vein networks to monitor how plants are responding to changing environments.

Details about the software, which can be downloaded for free, are reported in the journal Plant Physiology.

Leaf networks impact whole plant photosynthesis and the mechanical properties of leaves, and vary between species that have evolved or have been bred under different environmental conditions.

LEAF GUI provides structural measurements, including the dimensions, position, and connectivity of all network veins, and the dimensions, shape, and position of all non-vein areas, called areoles.

“The software can be used to help identify genes responsible for key leaf venation network traits and to test ecological and evolutionary hypotheses regarding the structure and function of leaf venation networks,” says Joshua Weitz, an assistant professor of biology at the Georgia Institute of Technology.

“The network extraction algorithms in LEAF GUI enable users with no technical expertise in image analysis to quantify the geometry of entire leaf networks—overcoming what was previously a difficult task due to the size and complexity of leaf venation patterns,” says the paper’s lead author Charles Price, who worked on the project as a postdoctoral fellow at Georgia Tech. Price is now an assistant professor of plant biology at the University of Western Australia.

While the Georgia Tech research team is currently using the software to extract network and areole information from leaves imaged under a wide range of conditions, LEAF GUI could also be used for other purposes, such as leaf classification and description.

“Because the software and the underlying code are freely available, other investigators have the option of modifying methods as necessary to answer specific questions or improve upon current approaches,” says Price.

The work was supported by the Defense Advanced Research Projects Agency, the National Science Foundation, and the Burroughs Welcome Fund.

More news from Georgia Tech: