摘要
在铝塑泡罩包装机的药片自动包装生产线上,药片板中往往会出现空泡、缺陷(缺片、斑点、裂片、黏片)等情况。针对此问题,本文利用VC++图像处理技术开发一套药片实时检测及处理的系统对药片板中的药片进行在线检测,通过改进型的中值滤波、差影法图像分割对药片板中的药片进行缺陷检测,再经过图像识别与分类方法把药片进行分类。本文在去除椒盐噪声方面比传统的方法有更好的滤波效果,且能很好解决药片与药片间的灰度交叉缺陷这一难点。该系统在实际的生产中对产品中缺粒、缺陷药片(缺片、斑点、裂片、黏片)等不合格的药品板准确检测率达到90%以上,满足了实际生产的需要。
On the automatic packaging lines which uses the aluminum-plastic blister packaging machine, in the tablets packaging process, tablet plate often appears empty, unqualified tablets（ such as unfilled tablet, impurities, fragments, sticky tablets）. Aiming at this problem, the paper develops a real - time tablets image detection and processing system by using the VC image processing technology. It is mainly used to detect unqualified tablets and tablet images detect with the improved median filtering, the difference image method, then classifies tablets with image identification and classification. This paper is more effect than traditional median filtering in getting ridof salt and pepper noise, it can resolve the defect of the gray cross between pills and pills, in the process of real production. It has detection ratio of 90% for tablet board which contains unqualified tablets（ such as unfilled tablet, impurities, fragments,sticky tablets）, meets the needs of actual production.
出处
《计算机与现代化》
2013年第5期66-69,共4页
Computer and Modernization
基金
国家自然科学基金资助项目(2001BA701A44A)
关键词
铝塑泡罩包装机
图像处理
中值滤波
aluminum-plastic blister packaging machine
image processing
median filtering
