A microscopic computer vision algorithm for autonomous bubble detection in aerated complex liquids. (December 2018)

• Record Type: Journal Article
• Authors: Misra, N.N.; Phalak, Rohit; Martynenko, Alex
• Is Part Of: Journal of food engineering. Volume 238(2018)
• Page Start: 54
• Page End: 60
• Publication Date: 2018-12
• Subjects: Cake batter -- Foam -- Aeration -- Image analysis -- Log-normal; Food industry and trade -- Periodicals; Food -- Analysis -- Periodicals; Aliments -- Industrie et commerce -- Périodiques; Aliments -- Analyse -- Périodiques; Aliments -- Recherche -- Périodiques; 664.005
• DOI: 10.1016/j.jfoodeng.2018.06.007
• Languages: English
• ISSNs: 0260-8774

Abstract

Abstract: The size distribution of bubbles in cake batters is often determined from optical microscopic imaging, considering the lesser availability and non-affordability of sophisticated techniques such as light scattering, and acoustic methods. We present an automated bubble detection and counting method from microscopic images that presents flexibility and robustness over existing manual approaches. The method is able to successfully resolve connected bubbles and recognise far many bubbles in an image than would be possible by naked eye or hitherto reported methods in chemical and food engineering literature. Furthermore, the size data obtained for the bubbles can easily be used for routine statistical analysis. We demonstrate the application of our method for studying the influence of two different mixer geometries and three different speeds on bubble size. Highlights: A direct microscopic imaging technique for bubble size evaluation in viscous liquids. A novel algorithm for bubbles quantification in noisy images. Method applied to study the effect of mixer speed and geometry on bubbles.