一种基于几何特征的多粒子粘连分割方法

在粉体加工和环境监测中,经常要测定粉尘的粒度及分布情况。用显微图像处理的方法进行粒度分析,技术先进而且能达到更好的效果,但对其中粘连重叠颗粒的分离,目前还没有很好的分割方法,针对这种情况本文提出了一种基于粒子的几何特征、采用平均链码方法求取凹点、并进行配对连线从而实现分割的方法。实验表明,该方法可以实现粘连颗粒的分离,并能达到很好的分割效果。

A Segmentation Method of Conglutination with Multi-particles Based on Its Geometrical Characteristics

In the case of powder processing and environment monitoring,it is common to detect the particle sizes and its distribution.The technology of microscopic image processing can be used in analysis of particles and better results will be achieved if the conglutination with multi-particles has been segmented perfectly.In this paper,a new segmentation method based on the geometrical characteristics of particles has been proposed.The particles are almost circular and there must be concave points when particles conglutinate together.The concave points here are allocated by average chain codes of the particle edges and the lines determined by the matched concave points separate the conglutinating particles into right parts.The results of experiment show that the method can realize the segmentation of the conglutinating particles and works well.