The paper presents a method for the macroscopic characterization of diesel sprays starting from digital images. Macroscopic spray characterization mainly consists in the definition of two parameters, namely penetration and cone angle. The latter can be evaluated according to many possible definitions, all based on the spray contour that is obtained by means of image thresholding. Therefore, the obtained cone angle value depends on the adopted angle definition and on the used thresholding algorithm. In order to avoid this double dependence, an alternative method has hence been proposed. The algorithm does not require the image thresholding and has an intrinsic cone angle definition. The algorithm takes advantage of principal component analysis technique and allows for a direct estimation of spray penetration and cone angle by comparing the original image with a database made of artificial spray images. In the present work, images coming from two different experiments are analyzed with the proposed method and results are compared with those obtained with a traditional procedure based on the Otsu's image thresholding and four cone angle definitions. © 2019 by the authors.
|Titolo:||Diesel spray macroscopic parameter estimation using a synthetic shapes database|
|Data di pubblicazione:||2019|
|Digital Object Identifier (DOI):||10.3390/app9235248|
|Appare nelle tipologie:||1.1 Articolo in rivista|