摘要
针对烟草企业对小包烟包装质量精密检测的需要,提出基于机器视觉的检测系列关键技术,包括了小包烟图像采集装置、基于快速Hough变换的小包边缘快速提取算法、基于灰度共生矩(GLCM)参数快速在线判别算法、基于支持向量机(SVMs)的包装缺陷识别等。该关键技术不仅适应在线快速包装质量检测,而且能准确判断缺陷类型。实验证明该关键技术具有实用可靠等优点,可应用到小包烟生产现场。
Aiming at the need of precise inspection of casing quality of small package in tobacco enterprise, a series of key inspection technologies was put forward, which was composed of image collecting device, edge detecting algorithm based on fast Hough transform, online differentiating algorithm based on GLCM parameters, casing flaws recognition algorithm based on SVMs. The key technologies not only are fit for online fast inspection, but can truly judge the flaws as well. Experiments showed that the inspection system is reliable and applicable, and could be applied in the manufacture field.
出处
《包装工程》
CAS
CSCD
北大核心
2007年第8期102-105,共4页
Packaging Engineering
作者简介
贺鑫（1970-），男，土家族，湖南湘西人，硕士，湖南对外经济贸易职业学院讲师，主要研究方向为包装机械及计算机教育管理。
