Combination of neural network and statistical methods for sensory evalution of biological products: On-line beauty selection of flowers

  • F. Ros
  • A. Brons
  • F. Sevila
  • G. Rabatel
  • C. Touzet
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 686)


In order to automize on-line selection of biological products, it is necessary to determine relationships between human sensory evaluation (like the beauty of flowerplants), and physical measurements on objects (like machine vision images). Classical methods of image processing and statistics, are combined with neural network techniques. The research deals with methods for the selection of significant parameters for the judgement, and methods for decision learning and generation: for both types of methods, classical statistics and neural network technics are either compared or combined. Interest of the various combinations are discussed, through the application on beauty selection of flowerplants.


Multilayer neural networks Principal Component Analysis Backward stepwise selection Sensory evaluation Image processing 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Journot V. 1991. Détermination d'un indice de la qualité des produits agricoles en frais. Thesis at the Institut National Agronomique, Paris-Grignon (in French).Google Scholar
  2. 2.
    Malek D.M., Munroe J.H., Schmitt D.J. and Korth B.. 1986. Statistical evaluation of sensory judges. Am. Soc. Brew. Chem. J.. 49(1):23.Google Scholar
  3. 3.
    Martens M. 1985. Sensory and chemical quality criteria for white cabbage studied by multivariate data analysis. Lebensmittel Wissenschaft und Techn. 18:100–104.Google Scholar
  4. 4.
    McDaniel M., Henderson L.A., Watson B.T.Jr. and Heatherball D.. 1987. Sensory panel training and screening for descriptive analysis of Pinot Noir wine fermented by several strains of malolactic bacteria. Journal Sensory Studies. 2:149.Google Scholar
  5. 5.
    Lippmann R.L.. 1987. An introduction to computing with neural nets. IEEE ASSP Magazine. Vol. 4. P4–22.Google Scholar
  6. 6.
    Pao Y. 1989. Adaptive Pattern Recognition and Neural Networks. Addison Wesley Pub. Comp.Google Scholar
  7. 7.
    Papelier S. 1990. Applications de la vision assistée par ordinateur à l'étude de la qualité du cyclamen. Report CNIH (in French).Google Scholar
  8. 8.
    Powers J.J. 1984. Using general statistical programs to evaluate sensory data. Food Technology 38(6):74–84.Google Scholar
  9. 9.
    Powers J.J. 1988. Uses of multivariate methods in screeening and training sensory panelists. Food Technology 42(11):123–127.Google Scholar
  10. 10.
    Resurreccion A.V.A. 1988. Applications of multivariate methods in food quality evaluation. 42(11):128–136.Google Scholar
  11. 11.
    Rumelhart, D.E., J.L.McClelland. 1986. Parallel Distributed Processing. MIT Press.Google Scholar
  12. 12.
    Slaughter, D.C., R.C. Harrell 1989. Discriminating fruit for robotic harvest using colour in natural outdoor scenes. TRANS. of the ASAE March–April 32(2):757–763.Google Scholar
  13. 13.
    Tillett R.D.. 1991. Image Analysis for Agricultural Processes: A Review of Potential Opportunities. J. Agricultural Engineering Research. Vol. 50(4): 247–258.Google Scholar
  14. 14.
    Tomassonne R. 1983. La régression linéaire: Nouveaux regards sur une ancienne méthode statistique. MASSON (in French).Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1993

Authors and Affiliations

  • F. Ros
    • 1
  • A. Brons
    • 1
  • F. Sevila
    • 1
  • G. Rabatel
    • 1
  • C. Touzet
    • 2
  1. 1.CEMAGREFMontpellier Cedex 1France
  2. 2.LERI-ERIEENîmesFrance

Personalised recommendations