DETECTION AND CLASSSICATION OF AIR BUBBLE SILHOUETTES IN PIVIMAGES

Marjo Lahti, Kalle Marjanen, Heimo Ihalainen

Abstract

A practical method for detection, segmentation and classification of the air bubble silhouettes in Particle Image Velocimeter (PIV) images is proposed. Air bubbles that are in focus plane have sharper edges and lower intensities than defocused bubbles. Detection and segmentation is based on estimating the local variances and comparing these with the local intensities. After detection and segmentation different statistics describing the size and shape of the bubbles can be measured. An example of using this method is provided.

Keywords

classification, air bubble, PIV

Download

PWC-2003-TC9-008.pdf

DOI

-

IMEKO TC

TC9 - Flow Measurement

Back to the proceedings