摘要
研究了针对烟条图像的图像处理、图像特征提取和图像识别的算法,设计了软件系统,完成了软件编程,并在生产现场进行了系统的安装、调试和实验。通过一系列的现场实验和测试,结果表明本系统可以对烟条外观品质进行有效的检测,并具有检测速度快、总体检测效果好、稳定性高的特点。
This thesis studies cigarette carton cases＇ image processing, image feature extraction and image recognition algorithm, designs software system and program and completes the system installation, debugging and the experiment in the work site at last. Through a series of site experiments and tests, the results show that this system can be used to effectively detect the cigarette carton cases quality and its has the advantages of fast inspection speed, good overall detection effect and high stability.
出处
《机械制造与自动化》
2015年第2期220-222,225,共4页
Machine Building & Automation
关键词
机器视觉
条烟外观
检测系统
图像识别
machine vision
cigarette carton
detection system
image recognition
