蔗糖结晶图像的特征提取与粒度分析  

Feature extraction and size analysis of sucrose particle image

作　　者：杨丹丹[1] 朱名日[1] 周光祥[1] 

机构地区：[1]桂林电子科技大学电子工程与自动化学院,广西桂林541004

出　　处：《桂林电子科技大学学报》2012年第2期155-161,共7页Journal of Guilin University of Electronic Technology

基　　金：广西教育厅科研基金(200808LX138)

摘　　要：针对目前蔗糖结晶过程仍然依靠人工目测粒度现状,研究了基于计算机视觉的蔗糖结晶颗粒的粒度分析方法。将分割后的蔗糖颗粒图像进行一系列后续处理,完整地分离出蔗糖结晶颗粒与背景,提取蔗糖结晶颗粒的粒度与形状特征参数后进行表面积当量直径的粒度分析。在粒度分布图形表征及其分布矩基础上,计算表征颗粒粒度的重要参数,如均匀度、平均孔径、变异系数等参数。实现用机器视觉代替人工视觉定量定性地分析与计算粒度,避免了人工目测中以经验判断粒度带来的误差。Nowadays towards granularity of sucrose crystallization process still relies on artificial visual,a sucrose particle size analysis method based on computer vision is presented.A complete separation of the sugar crystal particles and background is split visa a series of follow-up treatment towards image segmentation.Extract characteristic parameters about size and shape of the sucrose crystal particles,detecting particle size distribution of surface area equivalent diameter.Based on the graph characterization and the moment information of the crystal size distribution,calculating of the important parameters for characterization of particle size,such as uniformity,average pore diameter,coefficient of variation of parameters.Achieve machine vision instead of artificial vision,quantitative and qualitative analysis and calculation of particle size,to avoid the particle error caused by artificial visual experience.

关 键 词：蔗糖颗粒图像 特征提取 粒度分析 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]