›› 2012, Vol. 25 ›› Issue (12): 133-.

• 论文 • 上一篇    下一篇

机器视觉检测胶囊图像的预处理研究

郑瀚,韦文斌,齐子城   

  1. (1.广西工学院 电子信息与控制工程系,广西 柳州 545006;2.中国兵器科学研究院 测控技术与仪器研究所,浙江 宁波 310014)
  • 出版日期:2012-12-15 发布日期:2013-02-28
  • 作者简介:郑瀚(1987—),女,硕士研究生。研究方向:控制理论与控制工程。

Research on the Preliminary Treatment of Detected Capsule Image by Machine Vision

ZHENG Han, WEI Wen-Bin, QI Zi-Cheng   

  1. (1.Electronic Information and Control Engineering Department,Guangxi University of Technology,Liuzhou 545006,China;
    2.Control Technology and Instrument Research Institute,Academy of Science and Echnology of China Ordnance,Ningbo 310014,China)
  • Online:2012-12-15 Published:2013-02-28

摘要:

机器视觉检测是通过摄像头获取原始图像,但原始图像因检测特征不明显和噪声干扰等因素,不便直接用于识别。文中介绍了采用自适应中值滤波方法对图像进行滤波,并选择最大类间方差法对图像阈值进行分割。实验结果表明,采用这两种方法可以突出胶囊破裂的缺陷特征。

关键词: 机器视觉, 自适应中值滤波, 最大类间方差法

Abstract:

The detection by machine vision obtains original image via camera,however the low quality and noisy hindrance make it difficult to read the picture directly.This article adopts the adaptive-median-filter for image filtering,and the method of maximum variance for separating the defects of image.The experimental result shows that both methods are effective in detecting capsule defects.

Key words: machine vision;adaptive-median-filter;the method of maximum variance separate

中图分类号: 

  • TP391.41