密集粒子径迹的粘连分离

粒子径迹测量技术的核心问题为粘连径迹的分离。本文针对超密粒子径迹显微图像的粒子密度高、径迹粘连严重、颗粒度差异大、径迹内部灰度不均匀且存在杂质等特征，分析现有粘连分离算法存在的问题，提出了一种密集粒子径迹粘连分离方法。先对径迹显微图像进行增强，再进行扩展极大值分割，采用数学形态学及面积比例法分离图像内部粘连径迹，并对边缘处的径迹及无孔洞的单个径迹进行特殊处理。实验结果表明，算法能有效处理径迹显微图像，快速精确定位径迹轮廓，实现密集粒子径迹的粘连分离。

Overlapping Separation for Dense Particle Tracks

Overlapping separation of tracks is the core issue of particle tracks measurement. Aiming at the ultra-dense particle track microscopic image with features such as high density of particles, severe overlap of tracks, different level of granularity, uneven internal gray scale of tracks, and impurities existence, etc., the problems of the existing overlapping separation algorithms were analyzed, and a new method was proposed. The first step was enhancement for track microscopic image. Then the extended max segmentation was taken, and overlapping tracks were separated by mathematical morphology and area ratio method. At last, special process for tracks at the edge and single tracks with no hole was adopted. Experimental results show that the algorithm can process the track microscopic image effectively, locate the contours of tracks quickly and accurately, and separate dense overlapping particle tracks.