High-tech and Artificial Intelligence Make the Invisible Visible

Research and Development in Biosystems Engineering

LeafProcessor: Measuring and Analyzing Leaf Shape

The general function of leaves is supply a plant with energy and oxygen by means of photosynthesis. Leaves exist in many and diverse shapes that possibly represent differing evolutionary strategies of adapting to the changing environment. The specific shape of the leaves’ edges might be optimized for certain stress factors such as water deficiency. The shape of fossilized plant leaves are additionally used to estimate temperatures during their lifetimes (paleobotany). 

Studying the evolution of leaf shape delivers knowledge about the internal and external factors and mechanisms, which enable plants to produce leaves of a characteristic shape. A precise, quantizable description of leaf shapes is needed. Subtle changes caused by genetic modification, e.g. breeding, or selected growing conditions could also be evaluated.

LeafProcessor software was developed in collaboration with the Department of Animal and Plant Science, University of Sheffield, to give biologists and plant breeders a tool that evaluates different leaf geometries.

LeafProcessor Software Core Functions

  1. Organization: The software fully organizes image data generated in an experiment sorted in keeping with experimental conditions.
  2. Segmentation: Leaves in an image are segmented and marked automatically.
  3. Shape attributes: The software supplies a number of individual attributes with which shapes can be marked manually.
  4. Cluster analysis: Leaves can be automatically combined into groups with highly similar shapes.
  5. Classification: Based on the shape attributes, classification models can be generated and validated to automatically identify plant species or genotypes.
  6. Regression: Regressions models can be created to evaluate available quantitative response variables based on shape.

The software has been used by biologists to obtain important findings about the mechanisms of leaf growth, e.g. based on the distribution of cell division.

Die LEAFPROCESSOR Software zur Formbeschreibung und Formerkennung von Pflanzenblättern
© Fraunhofer IFF
LeafProcessor user interface. The software automatically marks shapes in an image and analyzes them with machine intelligence. Images can be marked in keeping with experimental conditions, which are then incorporated in report figures.


Backhaus, A.; Kuwabara, A.; Bauch, M.; Monk, N.; Sanguinetti, G. & Fleming, A., LEAFPROCESSOR: a new leaf phenotyping tool using contour bending energy and shape cluster analysis, New Phytologist, 2010, 187, 251-261

Backhaus, A.; Kuwabara, A.; Fleming, A. & Seiffert, U. Validation of Unsupervised Clustering Methods for Leaf Phenotype Screening, Proc. 18th European Symbosium on Artificial Neural Networks, Computational Intelligence and Machine Learning, 2010, 18, 511-517

Sloan, J.; Backhaus, A.; Malinowski, R.; McQueen-Mason, S. & Fleming, A. J. Phased control of expansin activity during leaf development identifies a sensitivity window for expansin-mediated induction of leaf growth, Plant Physiol., 2009, pp.109  

Kuwabara, A.; Backhaus, A.; Malinowski, R.; Bauch, M.; Hunt, L.; Nagata, T.; Monk, N.; Sanguinetti, G. & Fleming, A. A shift towards smaller cell size via manipulation of cell cycle gene expression acts to smoothen Arabidopsis leaf shape, Plant Physiology, 2011, 150, 2196-2206