软胶囊药板的机器视觉检测系统关键技术研究  

Study of Key Technology Used in Machine Vision Inspection System for Soft Capsule

作　　者：彭彦卿[1] 杨永通[1] 陈李清 

机构地区：[1]厦门理工学院电气工程与自动化学院,厦门361024 [2]国药控股星鲨制药(厦门)有限公司,厦门361026

出　　处：《化工与医药工程》2016年第2期39-42,共4页Chemical and Pharmaceutical Engineering

基　　金：厦门市科技计划项目(3502Z20141150)资助

摘　　要：基于机器视觉的软胶囊产品质量检测系统通过图像采集硬件将被检测软胶囊产品药板转换成图像信号,并传达给专用的图像处理系统。图像处理系统根据像素亮度、颜色分布等信息,抽取目标特征进行建模,建立缺陷图像数据库,通过图像处理和模式识别算法,判断药板是否存在质量问题,并将存在质量问题的药板从生产线上自动剔除。大量实验表明,检测准确率为99.7%以上,识别速率达2 400粒/min。In machine vision based soft capsule quality inspection system, by using collection hardware, soft capsule sheet inspected is transformed into image signals which are conveyed to image treatment system. In image treatment system, based on the information, such as brightness and color of images, the features of the objective are extracted and model establishment is finished. With these models the defected image database is established. By using the methods of image treatment and model recognition, capsule sheet is inspected if there is quality defect and unqualified capsule sheet can be automatically removed from production line. From the results of great number of experiment, it has been shown that using this system the precision rate may reach above 99.7% and speed rate of recognition may reach 2400 capsule/min.

关 键 词：机器视觉 软胶囊药板 图像采集 缺陷监测 

分 类 号：TQ460.7[化学工程—制药化工]