False positives: handling them operationally for automatic pollen monitoring
Benoît Crouzy, Gian-Duri Lieberherr, Fiona Tummon, Bernard Clot
Abstract
Abstract This note provides a generic method for handling falsely identified particles (false positive detections) in automatic pollen monitoring systems. The benefits of the method are illustrated using data from two different automatic pollen monitoring devices (Swisens Poleno and Plair Rapid-E) in comparison to reference Hirst-type pollen measurements. Neither climatological nor phenological information are introduced to filter out false positive detections: the method is based only on the measurements.
Topics & Concepts
PollenFalse positive paradoxComputer scienceFilter (signal processing)False positives and false negativesArtificial intelligenceComputer visionBiologyBotanyAllergic Rhinitis and SensitizationAdvanced Chemical Sensor TechnologiesInsect Pheromone Research and Control