![](bg4.png)
AgED provides three different views on the image
analysis pipeline: (i) The file system view lets the user
browse images in his file system. Furthermore, he can
label each image file with a class label. (ii) The shape
view allows for the inspection of each single shape
and associated EFDs that have been extracted from
the previously labeled image set. Note that one image
can contain several shapes. Also, the user can select
and deselect single shapes to be excluded from evalu-
ation. False positives and outliers can be removed this
way. (iii) In the evaluation view the probability den-
sities that have been calculated from the EFDs and its
Principal Components can be examined. In this view
the user can compare results and export the figures as
image files.
5 CONCLUSIONS
The results of this study showed that the automatic
and timely processing of high-resolution digital im-
age data for biological phenotype assessment appli-
cations is feasible. The AgED software toolkit fa-
cilitates and guides the preprocessing and allows the
browsing of the evaluation results. The applicabil-
ity of the AgED software tool has been exemplified
and validated on a real-world problem, that illustrates
some pitfalls diverse in severity. We have showed that
the AgED software deals effectively with these prob-
lems. While the presented image processing pipeline
readily scales to the automatic processing of a few
thousand images, the data evaluation step neverthe-
less demands some user interaction. We showed that
AgED reduces the time needed from the experimen-
tal setup to the deduction of the experiment’s result.
Furthermore, the extracted feature set as well as the
evaluation results are stored in a format that can be
adopted by other statistical software systems.
The software AgED can be downloaded from
http://sourceforge.net/projects/aged/. This repository
also comes with detailed instructions and examples.
Figure 4: Three thalli of Coleochaete scutata and their out-
lines (left). The according Fourier spectrum of these shapes
(right). The spectrum is displayed as fraction of the DC-
component. Irregular shapes convey larger magnitudes in
the Fourier spectrum than circular shapes.
REFERENCES
Chen, Y. and Nelson, R. L. (2004). Evaluation and Clas-
sification of Leaflet Shape and Size in Wild Soybean.
Crop Science, 44:671–677.
Costa, C., Antonucci, F., Boglione, C., Menesatti, P., Van-
deputte, M., and Chatain, B. (2012). Automated
sorting for size, sex, skeletal anomalies of cultured
seabass using external shape analysis. Aquacultural
Engineering.
Deelman, E., Gannon, D., Shields, M., and Taylor, I.
(2008). Workflows and e-Science: An Overview of
Workflow System Features and Capabilities. In Fu-
ture Generation Computer Systems. Elsevier.
Diaz, G., Zuccarelli, A., Pelligra, I., and Ghiani, A. (1989).
Elliptic Fourier Analysis of Cell and Nuclear Shapes.
Computers and Biomedical Research, 22:405–414.
Dupuy, L., Mackenzie, J., and Haseloff, J. (2010). Coordi-
nation of plant cell division and expansion in a simple
morphogenetic system. Proceedings of the National
Academy of Sciences, 107(6):2711–2716.
Ferson, S., Rohlf, F. J., and Koehn, R. K. (1985). Mea-
suring Shape Variation of Two-dimensional Outlines.
Systematic Biology, 34(1):59–68.
Iwata, H. and Ukai, Y. (2002). SHAPE: A Computer
Program Package for Quantitative Evaluation of Bio-
logical Shapes Based on Elliptic Fourier Descriptors.
Journal of Heredity, 93(5):384–385.
Kuhl, F. P. and Giardina, C. R. (1982). Elliptic Fourier Fea-
tures of a Closed Contour. Computer Graphics and
Image Processing, 18:236–258.
Laurie, C. C., True, J. R., Liu, J., and Mercer, J. M. (1997).
An Introgression Analysis of Quantitative Trait Loci
That Contribute to a Morphological Difference Be-
tween Drosophila simulans and D. mauritiana. Ge-
netics Society of America, 145:339–348.
Parzen, E. (1962). On Estimation of a Probability Density
Function and Mode. Annals of Mathematical Statis-
tics, 33(3):1065–1076.
Pryer, K. M. and Heam, D. J. (2009). Evolution of Leaf
Form in Marsileaceous Ferns: Ecidence for Hete-
rochrony. Evolution, pages 498–513.
Samet, H. and Tamminen, M. (1988). Efficient Component
Labeling of Images of Arbitrary Dimension Repre-
sented by Linear Bintrees. IEEE Transactions on Pat-
tern Analysis and Machine Intelligence, 10(4):579–
586.
Sprißler, N. (2007). Morphogenese von Coleochaete scu-
tata (Viridiplantae): Einfluss von Phytohormonen.
Dr. Burkhard Becker, Botanisches Institut, Universit
¨
at
K
¨
oln.
Truong, N. T., Gwag, J.-G., Park, Y.-J., and Lee, S.-H.
(2005). Genetic Diversity of Soybean Pod Shape
Based on Elliptic Fourier Descriptors. Korean Jour-
nal of Crop Science, 50(1):1–7.
Yoshioka, Y., Ohsawa, R., Iwata, H., Ninomiya, S., and
Fukuta, N. (2006). Quantitative Evaluation of Petal
Shape and Picotee Color Pattern in Lisianthus by Im-
age Analysis. Journal of the American Society for
Horticultural Science, 131(2):261–266.
AgED:ExtractionandEvaluationofEllipticFourierDescriptorsfromImageDatainPhenotypeAssessmentApplications
327