ii
Abstract
High quality atomization can improve combustion efficiency and reduce the emission of
pollutants. The macroscopic characteristic of spray is one of the key points of spray research.
The acquisition of macroscopic characteristic parameters is mainly to obtain images by
high-speed photography, and then to measure the macroscopic characteristic parameters by
image processing. In this regard, this paper focuses on the optimization of image processing
algorithms and the study of the method of parameter acquisition.
According to the actual characteristics of the spray oil beam image, a target image
segmentation algorithm is proposed. Compared with the existing algorithm, the result is better
than the commonly used threshold segmentation algorithm such as OTSU algorithm, which
can effectively protect the edge of spray oil beam. And the oil beam symmetry and
homogeneity are taken into account, this paper proposes the cone angle measurement method,
according to the analysis of the final rendering of the cone angle changes with the Naber
definition drawing cone angle contrast, found defined in this paper cone angle change trend
accords with Guang'an BoZhi spray cone angle formula, convergence and error and angle
fitting the curves are better than the Naber definition of cone angle curve. In order to keep the
original image's pixel value, we do not use the method of directly filtering the image to
remove noise. Instead, we use the method of obtaining the image boundary data first and then
filtering the noise. Median filter, mean filter and Kalman filter were selected respectively, and
the filtered data were analyzed through SPSS software. According to the analysis report, a
Kalman filter with higher regression and smaller error was selected.
Finally, based on Matlab, the GUI interface of fuel spray macro characteristic parameter
measurement is compiled, which is more convenient and intuitive to observe the measurement
process, save the measured data and analyze the measurement results.
Key words: Acquisition of spray image, Macro characteristics of spray, Measurement method
of spray angle, SPSS data analysis, Least-squares fitting of a straight line, Algorithm
optimization, Matlab/GUI
