| | |
Summary: AUTOMATIC BUBBLE NUCLEATION SITES IDENTIFICATION IN AN
IMAGE SEQUENCE
Da-chuan CHENG*, Hans BURKHARDT**
* Institute of Pattern Recognition and Image Processing, Department of Computer Science, University of
Freiburg, 79110 Freiburg, Germany, e-mail: cheng@informatik.uni-freiburg.de
** Institute of Pattern Recognition and Image Processing, Department of Computer Science, University of
Freiburg, 79110 Freiburg, Germany, e-mail: burkhardt@informatik.uni-freiburg.de
Abstract. In this study, we present an algorithm which can identify the nucleation sites of
vapour bubbles in an image sequence based on a template which is a bubble sub-image
extracted by the user. These images are taken with a speed of one thousand images per second
via a high-speed digital video camera. Our algorithm is divided into three parts: 1) the bubble
position detection, 2) small bubbles identification, and 3) nucleation sites identification via
tracking bubbles. We use backwards process in order to track bubbles.
The advantage of this method is to avoid the problem caused by bubble merging or the
identification of varying shapes of large bubbles. Afterwards, the positions of bubble
nucleation sites can be identified. It is an important parameter in exploring the heat transfer in
a boiling system.
1. Introduction. In order to explore the heat transfer relationship between the given liquids
and the heated tubes, the identification of the nucleation sites of vapor bubbles is important. The
density of the nucleation and its distribution area are some of important parameters to explore the
|
