1 / 20

A Semi-Automatic System for Pollen Recognition

A Semi-Automatic System for Pollen Recognition. Alain Boucher 1 , Régis Tomczak 2 , Pablo Hidalgo 3 , Monique Thonnat 1 , Jordina Belmonte 4 , Carmen Galan 3 and Pierre Bonton 2. 1- INRIA, Sophia-Antipolis, France 2- LASMEA, Blaise Pascal University, Clermont-Ferrand, France

chase-lee
Download Presentation

A Semi-Automatic System for Pollen Recognition

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. A Semi-Automatic System for Pollen Recognition Alain Boucher1, Régis Tomczak2, Pablo Hidalgo3, Monique Thonnat1, Jordina Belmonte4, Carmen Galan3andPierre Bonton2 1- INRIA, Sophia-Antipolis, France 2- LASMEA, Blaise Pascal University, Clermont-Ferrand, France 3- University of Córdoba (UCO), Spain 4- Autonomous University of Barcelona (UAB), Cerdanyola del Vallès,Spain

  2. Introduction • Goal: Help the technician identify and count per types the pollen grains • Pollen concentration can be used • for public report • for the forecast system

  3. System Architecture

  4. System Hardware A light microscope is driven automatically by a computer

  5. Material and Methods • System Hardware: • Light microscope • 3 axes micro-positionning device • CCD colour camera • Image acquisition card • PC computer • Pollen slides are prepared by technicians • Pollen grains are sampled using Hirst traps • Pollen grains are coloured with fuchsine • Pollen grains are observed with a magnification of 60x

  6. Man-Machine Interface The system can work in supervised or automatic mode

  7. Segmentation Localisation Pollen Grain Detection and Localisation Automatic pollen grain detection

  8. Pollen Grain Extraction • 3D acquisition of pollen grains • set of images at different depths • 100 optical sections • step = 0.5 microns For each grain Features may appear on different heights

  9. Main Pollen Types Studied and Similars Cupressaceae Olea Parietaria Poaceae Populus Brassicaceae Fraxinus Ligustrum Phillyrea Salix Broussonetia Morus Urtica membranacea Celtis Coriaria

  10. Pollen Grain Recognition • Goal is to identify the type of the pollen grain from 3D images • Grain recognition is done following two steps: • Compute global measures • Search for specific characteristics • Pollen knowledge is used to identify each grain • Palynology (apertures, reticulum, size, …) • Aerobiology (flowering period)

  11. Palynological Knowledge • The system tries to mimic the palynologists • Pollen knowledge is used to identify each grain • Knowledge sources from • Palynology • Aerobiology

  12. Help to decide which grain characteristics to look for Compute Global Measures Diameter (microns) vs Mean blue colour • Measures: • Diameter • Colour (RGB) • Shape • ...

  13. Example: Broken Cupressaceae Grains • Broken Cupressaceae grains are detected by shape: • Form factor: 4  surface / perimeter ² • Convexity ratio: grain surface / convex hull surface Broken grain Convex hull

  14. Search Pollen Characteristics • Only search for possible characteristics • Found characteristics help to look for others Blur analysis vs Image number Image Full Grain Inside Exine

  15. Example: Poaceae Pore Image 35 /100 Image 50 / 100 Image 65 / 100 Image 80 / 100

  16. Example: Cupressaceae Cytoplasm Image 40 / 100 Image 50 / 100 Image 60 / 100

  17. Example:Olea Reticulum Detection • The reticulum is located at top (or bottom) surface of the grain • Steps to follow: • Check if the pollen is reticulated • Localize the reticulum • Analyze the reticulum

  18. Example: Olea Reticulum Analysis • Find the image with the sharpest reticulum • Extraction of a zone with reticulum (muri & lumina) • Extraction of some lumina (clear or dark) • Analysis of size and shape of the lumina

  19. Pollen Counting • Final output: pollen concentration in the air • pollen count per type • hourly, dayly or weekly concentration • Unknown grains are reported to the technicians • Output can be used • for public report • for the forecast system

  20. Conclusion • Hard to give classification results so far • With only the 4 chosen pollen types, recognition of almost 100%, but need more tests including other similar pollen types • Database of more than 350 digitized pollen grains (30 different pollen types) • Steps of development : • 4 allergenic pollen types (Cupressaceae, Olea, Parietaria, Poaceae) • 11 similar pollen types (Populus, Fraxinus, Morus, Celtis, …) • 15 other frequent pollen types (Betula, Quercus, Pinus, …)

More Related