基于形状特征的液压传动和润滑系统固体污染物计算机视觉检测原理

Detection principle of computer vision based shape characteristic for solid contaminant in hydraulic transmission and lubricating system

传统的油液固体颗粒污染检测装置不能全面反映污染物特性，无法满足对系统进行在线监测、系统故障早期预报以及故障分析的要求。本文提出基于固体污染物形状特征的计算机视觉检测原理。可实现对固体污染物进行形状、大小、颜色的分类，从而有效地评价污染物的性质及可能的产生源，为预防故障、早期诊断故障提供了急需的信息。

This paper puts forward the detection principle of computer vision based shape characteristic of solid contaminant. The classification may be achieved for shape, size and colour off solid contaminant, thus with effectively evaluating the properties and sources of contaminants. The much-needed information is provided for the prevention and early diagnosis of faults.