文章基本信息

标题：IMAGE-BASED DETECTION OF NEEDLE-LIKE PARTICLE AGGLOMERATION WITH A CASE STUDY ON L-GLUTAMIC ACID CRYSTALLIZATION
本地全文：下载
作者：Yan Huo ; Fangkun Zhang
期刊名称：Fresenius Environmental Bulletin
印刷版ISSN：1018-4619
出版年度：2019
卷号：28
期号：12
页码：9452-9457
语种：English
出版社：PSP Publishing
摘要：Particle agglomeration commonly associated with crystallization processes has been increasingly studied for assessing the crystallization quality. An effective image analysis method is proposed for the agglomeration detection of needle-like particles during the crystallization process. The proposed method mainly consists of image pre-processing, primary sieving, re-segmentation, and particle classification. Firstly, the online captured images are efficiently pre-processed to reduce the influence from uneven illumination. Secondly, a primary sieving algorithm is established for discriminating candidate agglomerates, based on a shape feature. Thirdly, a re-segmentation algorithm is developed to extract the potential primary particles from the candidate agglomerates. In the end, a shape identification algorithm is given to recognize pseudo agglomerates from the candidate agglomerates based on two texture features. Thus, the agglomeration degree could be effectively assessed during the crystallization process. Experimental results on monitoring the cooling crystallization of β-form L-glutamic acid well demonstrate the effectiveness of the proposed image analysis method.
关键词：Particle agglomeration;crystallization detection;agglomerate identification;image processing;feature analysis