Criss cross : characterization and aggregation detection of needle shaped particles.

Abstract

In this project, the aim is to have an algorithm to automatically analyze sizes of crystals, inclusive of those overlapping. At the same time, it aims to exclude the aggregates from the analysis of the sizes of the crystals, and classify these aggregates into different classes of aggregation. Carbon fibre crystals are used during the analysis due to its elongated nature. The objective will be to have the algorithm report on the number and size distribution of crystals, inclusive of those overlapping, together with the number of aggregates and level of aggregation they are under in a very short time. The algorithm that had been developed gives an accuracy of around 99.9% for the measured length of particles. The time taken to analyze an image on average is 0.8 seconds.