Detection of bubbles as concentric circular arrangements

Abstract

The paper proposes a method for the detection of bubble-like transparent objects in a liquid. The detection problem is non-trivial since bubble appearance varies considerably due to different lighting conditions causing contrast reversal and multiple interreflections. We formulate the problem as the detection of concentric circular arrangements (CCA). The CCAs are recovered in a hypothesize-optimizeverify framework. The hypothesis generation is based on sampling from the partially linked components of the nonmaximum suppressed responses of oriented ridge filters, and is followed by the CCA parameter estimation. Parameter optimization is carried out by minimizing a novel cost-function. The performance was tested on gas dispersion images of pulp suspension and oil dispersion images. The mean error of gas/oil volume estimation was used as a performance criterion due to the fact that the main goal of the applications driving the research was the bubble volume estimation. The method achieved 28 and 13 % of gas and oil volume estimation errors correspondingly outperforming the OpenCV Circular Hough Transform in both cases and the WaldBoost detector in gas volume estimation.B Tuomas Eerola tuomas.eerola@lut.fi

From pattern recognition methods to machine vision applications

“…It would be interesting to know more about raw material at the pulping stage, also called the wet stage. In this research, a novel framework for bubble detection titled as concentric circular arrangements (CCA) was proposed [41]. In Figure 11.3, we show an example of gas volume estimation based on the detection of bubbles in pulp suspension.…”

Estimation of Bubble Size Distribution Based on Power Spectrum

“…1) that would be very difficult to do with sufficient quality. Therefore, as the reference method we use the bubble detection method based on concentric circular arrangements (CCA) [13] developed particularly for pulp suspension images. It has been shown to achieve good performance in the given task when the amount of bubbles is reasonable.…”

“…The typical approach to estimate the bubble size distribution is to first detect and segment the bubbles, and to compute the size of each detected bubble separately [13]. The bubble detection problem is not easy to solve because the bubbles are transparent and the illumination conditions are challenging when imaging inside an industrial process, which causes the bubble appearance to vary.…”

“…The geometry-based approaches suffer from a large number of false positives and are sensitive to noise. Moreover, they often fail to detect small blob-like bubbles that do not have a ridge edge expected by the model [13]. The appearance-based approach uses typically a sliding window where a template of the object of interest and the grayscale image are convolved.…”

Detection and tracking of bubbles in two‐phase air water flow for non‐convergent sinusoidal channel

“…Bubble detection techniques were further reviewed by Strokina et al . [12] separating the detection as geometric and appearance‐based approaches. In geometric approaches, a bubble model with circular or elliptical parameters is fitted on an image map.…”