Recognition Method for Engine Intake Gas Micelle State Based on Computed Tomography Image

In order to measure the density, shape and spatial distribution of engine intake gas micelles in real time, dynamic and non-destructive manner, this paper studies a recognition method for engine intake gas micelle state based on computed tomography image (CT image). First, the engine cross-section CT image is collected through the X-ray layer imaging system. Secondly, by extracting the cylinder circle, grayscale stretching and sub-pixel interpolation, the CT image of cylinder circle is obtained and its information is enhanced. Finally, based on the maximum inter-class variance, the characteristics of multi-target gas micelles have been measured, to obtain the engine intake gas clusters' characteristics of density, shape and spatial distribution. The results shows that the method can obtain engine intake gas clusters with distinct characteristics, and can combine various CT image chain structures to explain the gas micelles' movement pattern of each engine intake process, facilitate the analysis of engine intake state, improve engine thermal efficiency and environmental protection.