基于机器视觉的麻将包装识别系统设计
DOI:
CSTR:
作者:
作者单位:

(浙江工业大学 信息工程学院,杭州 310023)

作者简介:

胡文文(1991-),男,浙江宁波人,硕士研究生,主要从事嵌入式系统与应用方向的研究。 通讯作者:郑雅羽(1978-),男,浙江温州人,博士,副研究员,硕士生导师,主要从事视频编码和视频分析,嵌入式系统应用方向的研究。[FQ)]

通讯作者:

中图分类号:

基金项目:

国家自然科学基金资助项目(61401398)。


Mahjong Package Recognition System Design Based on Machine Vision[HS)]
Author:
Affiliation:

(College of Information Engineering, Zhejiang University of Technology, Hangzhou 310023, China)

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对麻将包装在实际产线上出现的麻将牌错位、缺失、牌面翻转等问题,基于机器视觉检测技术设计了一套以DM6437处理器为硬件平台的麻将包装识别处理系统;该系统的硬件部分由摄像头板、核心板和接口板3个模块组成,配合机械传动系统来进行识别前后麻将包装的传输;系统的软件部分采用平板Pad对麻将包装进行模板录入,并设计了一套适应于不同机箱内部环境变化的背景剔除方法来完成模板整体分割,以及通过中值计算实现单目标的分离,最后还提出了一种简单高效的混合搜索方式来进行基于SAD的模板匹配;测试结果表明,识别算法对于麻将间有缝隙、麻将整体倾斜、个别或整体麻将翻转、个别麻将位置错乱等问题均能准确判别,模板匹配算法稳定性和适应性较好,识别准确率达95%以上,能满足包装流水线上的缺陷检测。

    Abstract:

    To resolve the problem of Mahjong Card misplacement, deficit, and reversal for Mahjong packaging production line, a visual processing system of Mahjong package recognition is designed based on machine vision and DM6437 processor. The hardware system contains the sensor board, DSP board and interface board, and coordinates with mechanical transmission system to transport the Mahjong package before and after the recognition. Meanwhile, the Pad is used to achieve the Mahjong package templates entering in the software system, and a background subtraction method adaptive to the environment changes of different mechanical systems is designed to complete the templates global segmentation. Besides, the separation of single objective is realized by median algorithm. Eventually, a simple and efficient mixed searching method is presented to achieve the template matching based on SAD. Experiment results on the recognition system show that the algorithm has a strong tolerability on the gaps between Mahjong cards, the tilt and reversal of Mahjong packages, and the misplacement of Mahjong cards. Furthermore, the stability and adaptability of the template matching algorithm are quite effective, and the accuracy of the identification achieves more than 95%, which can meet the detection quality on the packaging lines.

    参考文献
    相似文献
    引证文献
引用本文

胡文文,王财盛,朱威,郑雅羽.基于机器视觉的麻将包装识别系统设计计算机测量与控制[J].,2016,24(5):258-262, 269.

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2015-11-06
  • 最后修改日期:2015-12-08
  • 录用日期:
  • 在线发布日期: 2016-07-27
  • 出版日期:
文章二维码