Erling Fløistad


Divisjon for bioteknologi og plantehelse

Divisjon for bioteknologi og plantehelse, stab

(+47) 970 89 021

Ås - Bygg H7

Høgskoleveien 7, 1433 Ås


Jobber med kommunikasjon, foto og bildearkiv

Les mer


This contribution demonstrates an example of experimental automatic image analysis to detect spores prepared on microscope slides derived from trapping. The application is to monitor aerial spore counts of the entomopathogenic fungus Pandora neoaphidis which may serve as a biological control agent for aphids. Automatic detection of such spores can therefore play a role in plant protection. The present approach for such detection is a modification of traditional manual microscopy of prepared slides, where autonomous image recording precedes computerised image analysis. The purpose of the present image analysis is to support human visual inspection of imagery data – not to replace it. The workflow has three components: • Preparation of slides for microscopy. • Image recording. • Computerised image processing where the initial part is, as usual, segmentation depending on the actual data product. Then comes identification of blobs, calculation of principal axes of blobs, symmetry operations and projection on a three parameter egg shape space.